Building high-ROI pipelines with agentic AI at Glean
Discover how Katy Abbaszadeh at Glean uses agentic AI to scale outcomes, boost pipeline ROI, and transform traditional marketing.
Sarah Casteel
No items found.
.jpg)
Discover how Katy Abbaszadeh at Glean uses agentic AI to scale outcomes, boost pipeline ROI, and transform traditional marketing.
This episode features an interview with Katy Abbaszadeh, Head of Pipeline Marketing at Glean, a leading work AI platform that connects company apps and data to help teams find information, collaborate, and make better decisions faster.
Katy shares how Glean is redefining pipeline generation with agentic AI, using autonomous agents to scale outcomes instead of just activities. She discusses how her team moved from “cute AI experiments” to leading with AI-first strategies that drive measurable impact, and what it takes to build a culture that embraces automation, experimentation, and continuous learning.
Key Takeaways:
TRANSCRIPT
Sarah McConnell – Qualified
Okay, Katy, thank you so much for joining us on The Agentic Marketer, where we're taking a peek into today's leaders’ tech stacks and AI strategies to learn how you're using AI agents to hit your pipe targets. But before we dive in, Katy, I'd love for you just to tell us a little bit about yourself and the work that you're doing over at Glean.
Katy Abbaszadeh – Glean
Sure, I'd be happy to. And first off, thanks for having me. Excited to be here. So I'm Katy Abbaszadeh. I'm responsible for what I refer to as all things pipeline marketing here at Glean. That includes demand gen, events, operations, really anything that's turning leads into meetings into pipeline. Yeah, go ahead.
Sarah McConnell – Qualified
Oh no, I love that. I'm thinking about my role, and I'm like, yeah, that is exactly what it is. Like all's I care about is pipeline.
Katy Abbaszadeh – Glean
Yep, that's the ultimate metric for us. And really the charter is high-quality pipeline at industry-leading ROI.
Sarah McConnell – Qualified
That's amazing. And now, obviously the whole show we're talking about agentic marketing. So obviously you're into ManGen, you're thinking about pipeline all the time, but I'm curious to kick things off. Like how do you define agentic marketing? Like what does that term mean to you and the team over at Glean?
Katy Abbaszadeh – Glean
Yeah, I think we're all at different stages in our AI and agentic roadmap, so to speak. So I'm so glad you have this forum for folks to share where they're at and what they're doing with it. I think for us, the future vision and where we're starting to get a foothold is that agentic marketing really means that we're orchestrating autonomous and semi-autonomous agents to do work for us. So it's about grounding the agents in your own data, but then making sure it's measured on things like revenue impact. We as marketers really care about the outcomes, not just the inputs. So making sure that the agents that you have working on your behalf have an outcome that they're working towards. So I think about it as not just implementing a set of new age tools, but really a new operating model for how we plan work, execute work, staff work.
Sarah McConnell – Qualified
Yeah, that's such a good answer. As I think about my next question is like, how is it different than traditional marketing? And at the, guess the base of it, seems pretty obvious, like obviously talking to AI and agents and that's very different. But do you feel like agentic marketing is vastly different than traditional marketing, or is it at its base the same thing we've always been doing just with like a different twist to it? But how do you think about those two things differently?
Katy Abbaszadeh – Glean
It's a good question. I think more recently I've come to believe that it is a bigger shift from traditional. When it was first coming out, I was like, that's cute that AI can do that. But now I'm more on the lines of thinking, this is about moving from in traditional marketing where we were scaling humans with tools. Now we're taking agents that can scale outcomes with systems. So thinking about moving from activities to that outcome base. You'll hear me talk about that a lot. It's focusing on what the end result is. Channels to agentic chains, different agents that can do different things, and fully autonomously is the ultimate vision. I'll admit we're not quite there yet, but there's a lot of promise there. And then from individual ops, QAing on a human level, to governance, systems of governance that can kind of be the guardrails to what your agents are doing.
Sarah McConnell – Qualified
Yeah, and I really like that answer because I feel like in 2023, when it was first like generative AI, that was like, oh, this is cute. Like it helps me. It's a little, we used to call the co-pilot. So you're riding shotgun and helping me make my job a little bit easier, but it just didn't feel like this massive shift. It was just like a slight change. But then now here in 2025, I agree. I'm like, oh no, this is vastly different. And we are being asked as marketers to have a completely different mindset and framework because we now do have these autonomous agents that we need to orchestrate and we need to have them do full roles that we are holding them accountable to an output, and that does feel vastly different than what we used to do. So I think your answer is perfect in that sense.
Katy Abbaszadeh – Glean
Yeah, 100 percent. I would say even—not to cut you off—but as we're heading into next year planning, you know, I'm starting to ask my managers, when you're thinking about headcount for next year, how does AI play into that? Not to the full extent of like we're freezing hiring until you prove that AI can't do this, but really thinking about AI as part of your team. And I haven't created a prototype prototype yet, but, you know, what does the new org chart look like, and how do you highlight the agents that are doing work?
Sarah McConnell – Qualified
That's a good segue to my next question, which I'm so curious. Making that leap from experimentation with AI to truly leading with it, I think in the way you just spoke about it, you are truly thinking about this as, 2026 planning, how is AI scaling your team? How are you thinking about bringing them on to help already existing org structures? So that is a big jump. So what did it take over at Glean for you to feel like you're making that move from the like, this is cute, this co-pilot is helping, to like, okay, no. This is a part of our 2026 planning. We need to lead with agentic AI and agents.
Katy Abbaszadeh – Glean
Yeah, it's a good question. And frankly, I don't think it's going to just be one single leap. I think it's actually baby steps to get to that ultimate leap. Any new skill, there's a learning curve. We're all adopting this new thing in real time. It's changing, feels like daily, weekly, new stuff popping on the scene. So I definitely think we have to give ourselves grace. And I try to think about leading with vulnerability. Not everything we test is going to come out the way we want it to. Not everything is going to be a smashing success. So how do we think about, as leaders, rolling up our sleeves, getting in the mix, being on the vendor calls, you know, workflow write-outs of jobs to be done, where can AI fit in? I think that's a really important piece of it. And I meant to introduce Glean lightly at the start, just because I think it helps to have some context of what Glean is. We're a work AI platform. So I have the benefit of working for an organization that is trying to transform work with AI across the board. So we're a horizontal platform that connects all of your apps and data at work in a safe way. So I have a bit of a cheat code in that my product that I'm marketing actually helps me be an AI-first marketer. But back to your question about what does that leap—I think the practical answer I'll give is I really recommend starting with one high-leverage workflow that you can really easily map out from what you're doing today and then kind of build in what are those things that can be replaced by AI. So an example is an ABM activation. Obviously, we hear a lot about content generation for AI being a major use case, personalization. So—
Sarah McConnell – Qualified
You're good.
Katy Abbaszadeh – Glean
—what is one ABM activation you're already doing that you can now kind of shift and think about it in an AI framework.
Sarah McConnell – Qualified
Yeah, I think that you talked about Glean being an AI-first company and giving you sort of that AI-first mindset, and obviously working at Qualified, I feel the same way. We have an AI-native product. We have our AI SDR, Piper, which we'll talk about, but it is such a unique position to be in because it does give you sort of the permission to really lead with that AI-first mentality, which I don't think is necessarily the case at every single organization. So I think to your point of starting there and it's starting with that experimentation, but picking like one thing, but being very lucky to work for organizations that do—
Katy Abbaszadeh – Glean
See you on the next video.
Sarah McConnell – Qualified
I'm assuming like your leadership team is asking you, like, where is AI showing up in your day-to-day because you're an AI product? So you're saying help to that accountability—
Katy Abbaszadeh – Glean
Yep. Yeah, absolutely. It takes forcing yourself to think about it every day. And I think at first, like we talked about it, it's so cute. It's fun. It's an experiment, right? But now, still maintaining time and space for the team to think about it, get creative with it, test and try and fail. I think the failures are really important in these baby steps, ultimate leap. So constantly trying to keep it top of mind.
Sarah McConnell – Qualified
Absolutely. Now, I'm curious, you said you oversee DemandGen, you oversee events and ops. Where is agentic AI showing up the most visibly on your team, whether that be from a DemandGen perspective or across your largest, like the marketing org as a whole? I think listeners are always so curious of, okay, agentic AI, agentic marketing sounds really cool, but I don't know where to put this into play. So where is it showing up for you on your Glean team most prominently?
Katy Abbaszadeh – Glean
Yeah, I think generally it's in those outcomes we talked about earlier, what we're kind of referring to as revenue moments. So you'll hear it come up when we get to the Piper part of the discussion, but that inbound speed to lead piece is really important to us. Outbound sequencing, some of those ABM moments where we can help the human SDRs that we do have be more personalized and proactive. Content generation and personalization we talked about already. Pre-meeting research for our reps that we as marketers can kind of curate for them. And then that, back to the inbound speed to lead, that 24/7 coverage that we just don't have to worry somebody's not going to get that custom touch that we want them to have. Obviously, productivity is a huge win across the board. So just planning time for campaign output.
Sarah McConnell – Qualified
Them.
Katy Abbaszadeh – Glean
You know, we've seen some of our planning cycles go from something like 15 days to a day because we can feed an agent the ideas, brainstorm with it, have a plan right then and there. And then more specifically within my team, I would say our digital marketing team is really leaning in heavily. So from a content generation perspective, we've got our SEO strategy really humming with AI workflows. Obviously, Piper is a big piece of that, our AI SDR on the website. And then even getting into some vibe coding on new website modules. So if you check out our homepage, you can see some that were vibe coded and then kind of built into the glean.com homepage. So some fun stuff there. And then the ABM stuff is a big play as well.
Sarah McConnell – Qualified
That's amazing. It's so cool to hear how many different aspects of your organization are adopting this, because I feel like getting it into one part of the organization always feels like that first step, and that can be not easy, but is easier to do, but getting it to be sort of an org-wide mindset, having been on the forefront of that, is much, much harder. And then my last question of this section of our conversation today, I'm very curious, do you have a hot take about AI and marketing that most people would not agree with you on?
Katy Abbaszadeh – Glean
Oof, hot takes. Love a hot take. I have to go back to the cutesy theme of how it burst onto the scene. I think right now most teams are overestimating the novelty and underestimating the operations portion of it, the backend portion of it. Agents are ready, and I'm not saying that Glean has it down perfectly just yet. They're still learning. But agents are ready, and it's really about—
Sarah McConnell – Qualified
Hahaha!
Sarah McConnell – Qualified
Them.
Katy Abbaszadeh – Glean
—org design, underlying data governance, how you're measuring things that become the bottlenecks to success. So I think the winners of this wave, so to speak, will really treat this as a go-to-market transformation and not just spot tools that you're rolling out here and there, but really rethinking how you plan, staff, and actually execute the work.
Sarah McConnell – Qualified
That's such a great answer because I do think it is a hot take because I do think we're hearing more in the market right now of people being like, well, maybe it's not as impactful as we thought it was going to be. And I worry that it's going to make marketers kind of rest on their laurels a little bit and be like, well, you know, if other people aren't finding success with it or I'm saying, you know, that's not impacting our bottom line, I don't want to invest the time and the effort to get up to speed with agentic AI and how I'm going to bring it onto my team. And I worry that with those two shifting narratives of like, no, it is really important versus like, OK, maybe there's not as much impact of it is that too many marketers will wait and say like, it isn't going to be that impactful. And then when it is, we'll feel left behind. So I love that your take is like, no, it's here. It's ready. It's kind of on us to build that infrastructure into our team. And I do, I agree with you. I think the teams that are embracing it early and are building in that infrastructure are the ones that are going to win and be way ahead. All of we're like, no, this is here to stay. This is going to be a vital part of our organizations. Encouraging other marketers to stay up to date with it and stay up to speed I think is so, so important.
Katy Abbaszadeh – Glean
Yeah. And I just want to say I understand the feeling of maybe burnout or, I tried this thing that I thought was going to be so great and I hit a wall and why would I try it again? I think even just having the ideas for, you know, certain workflows or what will be life changing when AI meets you there, having that scaffolding ready to go when AI catches up, you're that much further ahead. So. Keep it up.
Sarah McConnell – Qualified
Totally. Okay, so shifting more gears into what you guys are doing over at Glean, you guys have hired Piper, the AI SDR from Qualified, which is amazing. We're so happy to have you guys as customers. You've renamed her Taylor. What was it about having an AI SDR agent that made you and the team lean in? Like, what was it about that particular use case that the Glean team was like, okay, this is something I think we can start with or we can start utilizing within our organization?
Katy Abbaszadeh – Glean
Yeah, I think for us, as I mentioned at the top, all roads really lead back to PipeGen. It's the North Star metric for my entire team. Even Marketing Ops is constantly refreshing the dial on the Salesforce dashboard, even though they're not the ones with the exact ROI target. But the digital team in particular, who owns the website, really saw that pipeline was being left on the table because we couldn't rely on—
Sarah McConnell – Qualified
Mm-hmm.
Sarah McConnell – Qualified
What the—
Katy Abbaszadeh – Glean
—couldn't ask our human SDRs to be online 24/7. And so that team worked really closely with our Qualified CSM, Sean—shoutout—to put together a really compelling case to hire Piper. She is fully available on the website. We don't actually ask any of our human SDRs to be on that channel any longer. But I will say we don't plan on fully replacing our human SDRs. We really believe that there's an important place for them as well. But it was about pipeline. We didn't want to leave any dollar left behind.
Sarah McConnell – Qualified
I totally agree with you. And I feel like we went through the same sort of transformation here. I obviously get to be like on the forefront of our products and be our guinea pig, which is a blessing and a curse. We're the same way, and like asking our reps to be online all the time. Obviously one, can't—like they have to sleep, they have to eat. They cannot be online all the time. And even when they were, I just was just talking to someone else about tire deflection and like these, or sorry, tire kickers and like kind of deflection of people on the website is other, I'm like, am I really utilizing my SDR teams the right way when I'm seeing these conversations they're having with these people that aren't really that interested in like, yes, we can deflect them through the product. But I'm like, what if I—if I just have an AI agent do all of this, they can go focus on the things. And this is such a played out term, but like they can go focus on the things I want them to focus on and a lot more high-value things. Like cold outbounding is hard. It's a very hard thing to do. It's hard to do really well. We do have an outbound AI SDR that we use as well. But when it comes to like named accounts and really personalized experiences, I still very much believe in a human touch for that. So I'm like, if I can free up their time to do the good research and actually write really compelling messaging, is their time better spent doing that? So yeah, I agree with you on that. It's not about replacing them, but like, can we find areas for them that are gonna be more impactful for organization and for their career trajectory? Because responding to like mundane—
Katy Abbaszadeh – Glean
Thank you.
Sarah McConnell – Qualified
—answers on our website all day long, it's not really a great learning experience for them necessarily.
Katy Abbaszadeh – Glean
100%. And we still arm the human SDRs with their own AI for that personalization content generation. But I think another played-out term, human in the loop, is still really critical to success across the board.
Sarah McConnell – Qualified
Yeah.
Sarah McConnell – Qualified
And now beyond pipe generation, was there any unexpected things that bringing on an AI SDR agent helped unlock for your team? Is there anything you noticed that went beyond your expectations of pipeline generation and following up with leads?
Katy Abbaszadeh – Glean
Yeah, well, you actually just touched on it. And perhaps it shouldn't have been unexpected. But I do think the biggest eye-opener for me personally was the DQ rate. Piper does such a good job of weeding out prospects who just aren't a good fit, maybe yet or never will be. And so that saving for the human SDRs’ time was huge. So I think that was an eye-opener. And then just other metrics that we've been keeping a very close eye on with related to Piper is that time to first touch, you know, compared to the control of when the human SDRs were on when they could be, DQ rate, as I mentioned, and then meeting acceptance and show rate. Like, is there a difference between what Piper's able to book versus our SDR team? And then incremental meetings pipeline obviously is that outcome North Star that we're going for, and we've actually seen an uptick in that, I think due to the speed and efficiency, and she's always there.
Sarah McConnell – Qualified
That's incredible. You kind of already answered this, but I want to dig in a little bit deeper. As a marketing leader that's obviously taking AI very seriously within your organization, if you were talking to a peer and they're like, I just don't know where to start. I feel very overwhelmed. I think we've all felt that way at some point. What advice would you give to them on where to start? What is the best starting point to get over that hump of like, okay, I've run into too many walls with AI. I don't feel like this is worth my time. What advice would you give them?
Katy Abbaszadeh – Glean
This is a tough one because I think every organization is a little bit different, and depending on what your ultimate metric is, you want to make sure that your stake in the ground with AI has an impact on that. So I guess I would say, sort of back to the high-leverage use case, pick something that is a money-in, money-out workflow that you know like the back of your hand in the traditional way of marketing, and really try to map that out as specific jobs to be done and then rework that as an AI flow and do a comparison. Run a six-week A-B test and see what you get at the end. I think being clear about what you are measuring in that experiment is important, but then kind of let it do its thing and—what happens. I think that's the best way to really dip your toes in, and maybe in some cases traditional wins out because your underlying infrastructure isn't quite ready yet or, you know, I would say even if AI doesn't win the first battle, don't give up because there are always things that you can continue to refine and test.
Sarah McConnell – Qualified
And you said it earlier, I think it's okay to fail. I think that is something as marketers—don't, maybe it's just the circles that I run in—but marketers tend to be pretty type A for the most part. And I think failing is not something we love to do. I think it's a demand gen thing. I will say we tend to be cut from the same cloth. But failing is hard. Like I think if, as a demand gen person or someone that owns a pipeline number, you tend to be a little competitive. You've got big numbers that you're always looking to hit. You've got that drive in you, and failing, even if it is something that's new and different—
Katy Abbaszadeh – Glean
Guilty.
Katy Abbaszadeh – Glean
Yeah.
Sarah McConnell – Qualified
—can feel very frustrating. So I like that you said that early about be okay failing, but if you set the outcomes that you're looking for, it makes it a lot easier to fail fast and say like, okay, this isn't working, let's pivot, let's move on versus just spinning your wheels on something for a long time that ends up not playing out and/or giving up too fast. Like it might be something that you want to stick to. So I think your advice is great and also like fail quick and move on.
Katy Abbaszadeh – Glean
Yeah, I was just at an AI for Women meetup in Minneapolis last week, moderated a panel there. And I think that was my number one takeaway. It was really, we have to be able to embrace failure right now. And this wasn't marketing specific. It was kind of across the board—sales enablement, IT folks. But it's so new and changing so fast that if you expect it to be perfect and intuitive and something you can just roll out and walk away from—
Sarah McConnell – Qualified
Yeah.
Katy Abbaszadeh – Glean
—then you're definitely setting yourself up for failure. But if you take the bite-size failures along the way, like you said, fail fast and learn from it, then you're setting yourselves up—and your team—for success.
Sarah McConnell – Qualified
Absolutely. Now, as we think about AI in general, it's obviously made a massive impact on us as marketers and within our organizations. But I think it's also making a pretty big impact on the expectations of buyers. So do you have any advice for us as marketers? Like, how do you think AI is changing buyer expectations? And are we underestimating anything that we need to be cognizant of that AI is really changing for our buyers that we need to keep up with?
Katy Abbaszadeh – Glean
I'm gonna steal an answer from a CEO of a vendor I chatted with a few months ago, but it has just stuck with me. Eventually, you can imagine a world where we're marketing not to humans anymore, but to agents who bring back answers from queries on ChatGBT. Those agents are gonna go out to the websites, figure out how to compare different products and—
Sarah McConnell – Qualified
Mmm.
Katy Abbaszadeh – Glean
—serve up an answer where buyers expect a one-minute internet immediate answer—what's the best choice? And so I think we as marketers really need to be aware that some of our key metrics are going to shift. Should we care as much about website visitors in the future? Because it might just be an agent pinging your website or having it cached. So I don't know. It's a bit of a scary frontier because we're—
Sarah McConnell – Qualified
Yeah.
Katy Abbaszadeh – Glean
—we love our websites, don't we? But I do think we have to think about how our metrics shift, who we're marketing to shifts. So a lot to think about on that front.
Sarah McConnell – Qualified
I just had someone on the podcast last week and they kind of shared the same sentiment. And I think that's been like prolific around the marketing—just any peers I've talked to is like, yeah, website traffic is down. Like I think LLMs and the way that people are engaging with search is just changing so drastically, so quickly. But the continued sentiment has been the traffic that is coming from those, the AI sources, tends to be a much higher quality because they've already furthered themselves so much farther down the buying cycle, because of the way they're engaging with them. And at some point it's gonna be agents coming in, querying your website, and they're gonna come back with an answer as the buying cycles go faster. There are much higher conversion rates of those visitors coming from those sources converting into opportunities, and then once they're opportunities, they're closing quicker. As marketers, it will be so—it'll be a new challenge to convince our executive teams that like, yeah, it's okay that website traffic is dropping because the outputs and the metrics that we care about are still—as long as you're maintaining those—it's fine. But it is very scary to suddenly see this dramatic drop in traffic and sources. That's something I think we've all been struggling with just based on the conversations that I've been having.
Katy Abbaszadeh – Glean
Yeah, totally agree with that. And while the traffic from the LLMs is still small compared to everything else, to your point, the quality and the conversion and the speed is so much better. I think we just have to continue focusing on outcomes and that ultimate outcome that we're looking for versus the input metrics that we've typically kind of hung our hats on. So I think it makes it that much more important to drive that alignment with sales. What is the definition of pipeline? At what point, at what stage in the cycle are we measuring it? What coverage do we need? And making sure that we as marketers are speaking that same language so that we're not over here saying, the traffic, but the MQL.
Sarah McConnell – Qualified
Hahaha. Now, on that same note, like we just talked about, obviously search and web traffic has been a huge change, but is there another big—like, what would you say is gonna be the single biggest change to B2B marketing in the next 12 months, or is it going to be search in the way that people are interacting with our website?
Katy Abbaszadeh – Glean
Yeah, predicting the future. Yes, here we go. You know, I think this is less in my wheelhouse from a demand perspective, but as a buyer myself, I'm feeling like this shift is something we can't ignore. Pricing structure and pricing expectations. People do not want to get locked into—
Sarah McConnell – Qualified
I know, I'm like, hey, can you get out your crystal ball and just like tell me what the future holds, please?
Sarah McConnell – Qualified
Mm-hmm.
Sarah McConnell – Qualified
Yeah.
Katy Abbaszadeh – Glean
—annual contracts because things are changing so fast and you want to be able to test. So I think it's maybe a little bit easier to price that way in a consumer market, but on the B2B side, we have to be thinking about how we shift to that expectation so that people can be confident, you know, testing and using a product without getting locked in for one- to two-year contracts.
Sarah McConnell – Qualified
Yeah, I think that's a really good one. And earlier we've kind of talked—I think we've touched a lot throughout this conversation on use cases for AI agents. Like, I think you gave some really great examples that you're seeing within your own marketing organization. But one of the things I'm so curious about is what do you hope to see more of soon? Like, are there any areas within your marketing team that you're like, man, I really wish there was some sort of agentic solution that could help me scale this that you're just not seeing as much of in the market right now?
Katy Abbaszadeh – Glean
Yeah, this one's maybe less marketing focused, but I really am liking the agents that are more focused on coaching. So, executive coaching—I'm, you know, moving up in my career. How can I make sure I'm speaking to the right audience and the right tone and saying the right things? For our sales team, first pitch coaching. So I think that will probably become more formalized. It's a really interesting use case. One that's not necessarily a use case, but I do think we need to get there as an industry is open metrics standards, so that we can actually compare agent performance apples to apples. Obviously, each company is going to have a tech stack, different way of building agents. But if there's a way that we can—
Sarah McConnell – Qualified
Hmm.
Katy Abbaszadeh – Glean
—as marketers, always asking, OK, what's the industry average for CAC and web visits and all of that? We need a similar way to benchmark how we're doing in AI, and I don't think we're quite there yet.
Sarah McConnell – Qualified
That's probably my favorite answer I've heard today. And as you said, industry benchmarks, I like had a shiver run through me because I think in demand gen, the number of times we're like, well, what's the industry after—I'm like, I don't, I don't know. Like it's so different across organizations. Like in every company I've worked at, it's been so different. I tried to crowdsource recently, like as website traffic has been dropping, I'm like, where are people seeing—like what's your percentage breakdowns of web traffic? And it was—when I say all over the board is an understatement. So being able to have more standardized averages, especially as we think about AI, like, I like jump for joy. I was like, yes, can someone please make this if you're listening, because we just don't have that insight. And I feel like we're always flying a little bit blind.
Katy Abbaszadeh – Glean
Yeah, well hopefully the LLMs can already help because they've got access to all of the information. So you can get more specific on what do we even mean by industry when we say industry average. So I think that's promising, but we gotta do some work to make sure it's there for everyone.
Sarah McConnell – Qualified
Totally. Okay, my last question before we get into my favorite section, which is the lightning round. I'm curious if there is something that you think most B2B orgs will still get wrong about AI a year from now. So before I ask you—you know, get out your crystal ball, pull it back out—what do you think we're getting wrong now that you don't think is gonna get better? Or that, I guess I want to phrase that is, what do you think we should be focusing on because you're worried it's a problem or that we're still getting wrong?
Katy Abbaszadeh – Glean
Yeah, I'll do a callback to one that I kind of mentioned earlier, but I think it's about focusing too much or celebrating too much the demoable moments, the like little delightful assistance that like, hey, look how cute this is, and ignoring what that boring backbone takes to make it work more in a complex manner. So the data governance—at that breakfast I was at last week, there was a ton of talk about changing the terminology of data governance to AI governance. And how do we—you know, another phrase we like to use, garbage in, garbage out—how do we make sure that the AI is set up for success? And building that foundation is hard. And there's a lot of question on who owns building that foundation. You know, when you're in marketing and—
Sarah McConnell – Qualified
Mm-hmm.
Katy Abbaszadeh – Glean
—you're partnering with rev ops or IT, you know, who really owns that and how do you make sure that's set up appropriately? I think that'll be the tough thing for people to get right. So I'm interested to see how we solve that problem collectively.
Sarah McConnell – Qualified
Yeah, that is a great answer. Okay, I love to wrap up these episodes with our lightning round. So it's quick questions with quick answers. So to get started, other than ChatGPT, what was one of the first AI tools that you experimented with?
Katy Abbaszadeh – Glean
This is going to sound really lame. I promise I'm not going to sell Glean. But it was Glean because I was just coming back from maternity leave right after ChatGBT exploded onto the scene. And our engineering team had just done a hackathon. And three weeks after the hackathon, we launched Glean Chat. And so I really played with ChatGBT through Glean before I even played directly with ChatGBT.
Sarah McConnell – Qualified
That is such a—what a wild thing to come back to maternity leave from is like, I get back from maternity leave in August and I was like, what is vibe coding? Like, can someone—that's been four months and I feel like suddenly it was a completely different world. And that was the least dirty day. I feel like you came back to like a whole—literally a whole new world. Okay, what do you think is the most overrated buzzword in martech right now?
Katy Abbaszadeh – Glean
Right? Exact same experience. Yes, it was insane. I'm gonna go with personalization at scale. And I think this one is one that we like to use. It sounds great. We all talk about ABM. But without that governance piece and a measurement on outcome, it's just louder spray and pray. So.
Sarah McConnell – Qualified
It's so true. Okay, who is a marketer that you think is ahead of the curve on AI that listeners should go follow on LinkedIn right now?
Katy Abbaszadeh – Glean
OK, I will give you a name, but I first want to start with a shoutout to another podcast that I recently found. So on top of this podcast, which obviously you're already following because you're listening to this, Humans of MarTech has been a really great one that I've recently discovered. So shoutout to them, curating a ton of really interesting content from marketing leaders. And then, for an individual person, I have to give a shoutout to our head of digital marketing, Daniel Henderson. He has been partnering closely with Kevin Indig, who is somewhat of an AISCO influencer out there. And he's been involved in a lot of our AISCO projects. He's got the scrappy, rapid testing approach that has really helped us kind of frame our thinking and where we want to place bets. So Kevin Indig is a good one to check out.
Sarah McConnell – Qualified
Amazing. Okay, last question. If you could automate one part of your life with AI that's outside of work, what would it be?
Katy Abbaszadeh – Glean
I would absolutely love an AI stylist or personal shopper. And I'm sure there's a version of this that probably already exists, or I could go train my own custom GPT, but who has the time for that?
Sarah McConnell – Qualified
Absolutely. So I agree with that. Okay, Katy, I appreciate you joining us so much. Thank you for joining The Agentic Marketer podcast. It was great having you on and having you your insights. So thank you so much.
Katy Abbaszadeh – Glean
Thank you, Sarah.
.avif)
Discover how ROI·DNA’s Gina Inks uses agentic AI to scale B2B marketing, boost pipeline performance, and build lasting buyer trust in a zero-click world.
.avif)
Learn how CMO William Tyree leverages agentic marketing and an AI SDR to improve lead response, refine buyer experiences, and drive pipeline growth.
Discover how LogicMonitor CMO Brooke Cunningham is using AI-first marketing and agentic strategies to accelerate pipeline, boost organic traffic, and shape the future of B2B demand generation.
Stay up to date with weekly drops of fresh B2B marketing and sales content.
Discover how Katy Abbaszadeh at Glean uses agentic AI to scale outcomes, boost pipeline ROI, and transform traditional marketing.
This episode features an interview with Katy Abbaszadeh, Head of Pipeline Marketing at Glean, a leading work AI platform that connects company apps and data to help teams find information, collaborate, and make better decisions faster.
Katy shares how Glean is redefining pipeline generation with agentic AI, using autonomous agents to scale outcomes instead of just activities. She discusses how her team moved from “cute AI experiments” to leading with AI-first strategies that drive measurable impact, and what it takes to build a culture that embraces automation, experimentation, and continuous learning.
Key Takeaways:
TRANSCRIPT
Sarah McConnell – Qualified
Okay, Katy, thank you so much for joining us on The Agentic Marketer, where we're taking a peek into today's leaders’ tech stacks and AI strategies to learn how you're using AI agents to hit your pipe targets. But before we dive in, Katy, I'd love for you just to tell us a little bit about yourself and the work that you're doing over at Glean.
Katy Abbaszadeh – Glean
Sure, I'd be happy to. And first off, thanks for having me. Excited to be here. So I'm Katy Abbaszadeh. I'm responsible for what I refer to as all things pipeline marketing here at Glean. That includes demand gen, events, operations, really anything that's turning leads into meetings into pipeline. Yeah, go ahead.
Sarah McConnell – Qualified
Oh no, I love that. I'm thinking about my role, and I'm like, yeah, that is exactly what it is. Like all's I care about is pipeline.
Katy Abbaszadeh – Glean
Yep, that's the ultimate metric for us. And really the charter is high-quality pipeline at industry-leading ROI.
Sarah McConnell – Qualified
That's amazing. And now, obviously the whole show we're talking about agentic marketing. So obviously you're into ManGen, you're thinking about pipeline all the time, but I'm curious to kick things off. Like how do you define agentic marketing? Like what does that term mean to you and the team over at Glean?
Katy Abbaszadeh – Glean
Yeah, I think we're all at different stages in our AI and agentic roadmap, so to speak. So I'm so glad you have this forum for folks to share where they're at and what they're doing with it. I think for us, the future vision and where we're starting to get a foothold is that agentic marketing really means that we're orchestrating autonomous and semi-autonomous agents to do work for us. So it's about grounding the agents in your own data, but then making sure it's measured on things like revenue impact. We as marketers really care about the outcomes, not just the inputs. So making sure that the agents that you have working on your behalf have an outcome that they're working towards. So I think about it as not just implementing a set of new age tools, but really a new operating model for how we plan work, execute work, staff work.
Sarah McConnell – Qualified
Yeah, that's such a good answer. As I think about my next question is like, how is it different than traditional marketing? And at the, guess the base of it, seems pretty obvious, like obviously talking to AI and agents and that's very different. But do you feel like agentic marketing is vastly different than traditional marketing, or is it at its base the same thing we've always been doing just with like a different twist to it? But how do you think about those two things differently?
Katy Abbaszadeh – Glean
It's a good question. I think more recently I've come to believe that it is a bigger shift from traditional. When it was first coming out, I was like, that's cute that AI can do that. But now I'm more on the lines of thinking, this is about moving from in traditional marketing where we were scaling humans with tools. Now we're taking agents that can scale outcomes with systems. So thinking about moving from activities to that outcome base. You'll hear me talk about that a lot. It's focusing on what the end result is. Channels to agentic chains, different agents that can do different things, and fully autonomously is the ultimate vision. I'll admit we're not quite there yet, but there's a lot of promise there. And then from individual ops, QAing on a human level, to governance, systems of governance that can kind of be the guardrails to what your agents are doing.
Sarah McConnell – Qualified
Yeah, and I really like that answer because I feel like in 2023, when it was first like generative AI, that was like, oh, this is cute. Like it helps me. It's a little, we used to call the co-pilot. So you're riding shotgun and helping me make my job a little bit easier, but it just didn't feel like this massive shift. It was just like a slight change. But then now here in 2025, I agree. I'm like, oh no, this is vastly different. And we are being asked as marketers to have a completely different mindset and framework because we now do have these autonomous agents that we need to orchestrate and we need to have them do full roles that we are holding them accountable to an output, and that does feel vastly different than what we used to do. So I think your answer is perfect in that sense.
Katy Abbaszadeh – Glean
Yeah, 100 percent. I would say even—not to cut you off—but as we're heading into next year planning, you know, I'm starting to ask my managers, when you're thinking about headcount for next year, how does AI play into that? Not to the full extent of like we're freezing hiring until you prove that AI can't do this, but really thinking about AI as part of your team. And I haven't created a prototype prototype yet, but, you know, what does the new org chart look like, and how do you highlight the agents that are doing work?
Sarah McConnell – Qualified
That's a good segue to my next question, which I'm so curious. Making that leap from experimentation with AI to truly leading with it, I think in the way you just spoke about it, you are truly thinking about this as, 2026 planning, how is AI scaling your team? How are you thinking about bringing them on to help already existing org structures? So that is a big jump. So what did it take over at Glean for you to feel like you're making that move from the like, this is cute, this co-pilot is helping, to like, okay, no. This is a part of our 2026 planning. We need to lead with agentic AI and agents.
Katy Abbaszadeh – Glean
Yeah, it's a good question. And frankly, I don't think it's going to just be one single leap. I think it's actually baby steps to get to that ultimate leap. Any new skill, there's a learning curve. We're all adopting this new thing in real time. It's changing, feels like daily, weekly, new stuff popping on the scene. So I definitely think we have to give ourselves grace. And I try to think about leading with vulnerability. Not everything we test is going to come out the way we want it to. Not everything is going to be a smashing success. So how do we think about, as leaders, rolling up our sleeves, getting in the mix, being on the vendor calls, you know, workflow write-outs of jobs to be done, where can AI fit in? I think that's a really important piece of it. And I meant to introduce Glean lightly at the start, just because I think it helps to have some context of what Glean is. We're a work AI platform. So I have the benefit of working for an organization that is trying to transform work with AI across the board. So we're a horizontal platform that connects all of your apps and data at work in a safe way. So I have a bit of a cheat code in that my product that I'm marketing actually helps me be an AI-first marketer. But back to your question about what does that leap—I think the practical answer I'll give is I really recommend starting with one high-leverage workflow that you can really easily map out from what you're doing today and then kind of build in what are those things that can be replaced by AI. So an example is an ABM activation. Obviously, we hear a lot about content generation for AI being a major use case, personalization. So—
Sarah McConnell – Qualified
You're good.
Katy Abbaszadeh – Glean
—what is one ABM activation you're already doing that you can now kind of shift and think about it in an AI framework.
Sarah McConnell – Qualified
Yeah, I think that you talked about Glean being an AI-first company and giving you sort of that AI-first mindset, and obviously working at Qualified, I feel the same way. We have an AI-native product. We have our AI SDR, Piper, which we'll talk about, but it is such a unique position to be in because it does give you sort of the permission to really lead with that AI-first mentality, which I don't think is necessarily the case at every single organization. So I think to your point of starting there and it's starting with that experimentation, but picking like one thing, but being very lucky to work for organizations that do—
Katy Abbaszadeh – Glean
See you on the next video.
Sarah McConnell – Qualified
I'm assuming like your leadership team is asking you, like, where is AI showing up in your day-to-day because you're an AI product? So you're saying help to that accountability—
Katy Abbaszadeh – Glean
Yep. Yeah, absolutely. It takes forcing yourself to think about it every day. And I think at first, like we talked about it, it's so cute. It's fun. It's an experiment, right? But now, still maintaining time and space for the team to think about it, get creative with it, test and try and fail. I think the failures are really important in these baby steps, ultimate leap. So constantly trying to keep it top of mind.
Sarah McConnell – Qualified
Absolutely. Now, I'm curious, you said you oversee DemandGen, you oversee events and ops. Where is agentic AI showing up the most visibly on your team, whether that be from a DemandGen perspective or across your largest, like the marketing org as a whole? I think listeners are always so curious of, okay, agentic AI, agentic marketing sounds really cool, but I don't know where to put this into play. So where is it showing up for you on your Glean team most prominently?
Katy Abbaszadeh – Glean
Yeah, I think generally it's in those outcomes we talked about earlier, what we're kind of referring to as revenue moments. So you'll hear it come up when we get to the Piper part of the discussion, but that inbound speed to lead piece is really important to us. Outbound sequencing, some of those ABM moments where we can help the human SDRs that we do have be more personalized and proactive. Content generation and personalization we talked about already. Pre-meeting research for our reps that we as marketers can kind of curate for them. And then that, back to the inbound speed to lead, that 24/7 coverage that we just don't have to worry somebody's not going to get that custom touch that we want them to have. Obviously, productivity is a huge win across the board. So just planning time for campaign output.
Sarah McConnell – Qualified
Them.
Katy Abbaszadeh – Glean
You know, we've seen some of our planning cycles go from something like 15 days to a day because we can feed an agent the ideas, brainstorm with it, have a plan right then and there. And then more specifically within my team, I would say our digital marketing team is really leaning in heavily. So from a content generation perspective, we've got our SEO strategy really humming with AI workflows. Obviously, Piper is a big piece of that, our AI SDR on the website. And then even getting into some vibe coding on new website modules. So if you check out our homepage, you can see some that were vibe coded and then kind of built into the glean.com homepage. So some fun stuff there. And then the ABM stuff is a big play as well.
Sarah McConnell – Qualified
That's amazing. It's so cool to hear how many different aspects of your organization are adopting this, because I feel like getting it into one part of the organization always feels like that first step, and that can be not easy, but is easier to do, but getting it to be sort of an org-wide mindset, having been on the forefront of that, is much, much harder. And then my last question of this section of our conversation today, I'm very curious, do you have a hot take about AI and marketing that most people would not agree with you on?
Katy Abbaszadeh – Glean
Oof, hot takes. Love a hot take. I have to go back to the cutesy theme of how it burst onto the scene. I think right now most teams are overestimating the novelty and underestimating the operations portion of it, the backend portion of it. Agents are ready, and I'm not saying that Glean has it down perfectly just yet. They're still learning. But agents are ready, and it's really about—
Sarah McConnell – Qualified
Hahaha!
Sarah McConnell – Qualified
Them.
Katy Abbaszadeh – Glean
—org design, underlying data governance, how you're measuring things that become the bottlenecks to success. So I think the winners of this wave, so to speak, will really treat this as a go-to-market transformation and not just spot tools that you're rolling out here and there, but really rethinking how you plan, staff, and actually execute the work.
Sarah McConnell – Qualified
That's such a great answer because I do think it is a hot take because I do think we're hearing more in the market right now of people being like, well, maybe it's not as impactful as we thought it was going to be. And I worry that it's going to make marketers kind of rest on their laurels a little bit and be like, well, you know, if other people aren't finding success with it or I'm saying, you know, that's not impacting our bottom line, I don't want to invest the time and the effort to get up to speed with agentic AI and how I'm going to bring it onto my team. And I worry that with those two shifting narratives of like, no, it is really important versus like, OK, maybe there's not as much impact of it is that too many marketers will wait and say like, it isn't going to be that impactful. And then when it is, we'll feel left behind. So I love that your take is like, no, it's here. It's ready. It's kind of on us to build that infrastructure into our team. And I do, I agree with you. I think the teams that are embracing it early and are building in that infrastructure are the ones that are going to win and be way ahead. All of we're like, no, this is here to stay. This is going to be a vital part of our organizations. Encouraging other marketers to stay up to date with it and stay up to speed I think is so, so important.
Katy Abbaszadeh – Glean
Yeah. And I just want to say I understand the feeling of maybe burnout or, I tried this thing that I thought was going to be so great and I hit a wall and why would I try it again? I think even just having the ideas for, you know, certain workflows or what will be life changing when AI meets you there, having that scaffolding ready to go when AI catches up, you're that much further ahead. So. Keep it up.
Sarah McConnell – Qualified
Totally. Okay, so shifting more gears into what you guys are doing over at Glean, you guys have hired Piper, the AI SDR from Qualified, which is amazing. We're so happy to have you guys as customers. You've renamed her Taylor. What was it about having an AI SDR agent that made you and the team lean in? Like, what was it about that particular use case that the Glean team was like, okay, this is something I think we can start with or we can start utilizing within our organization?
Katy Abbaszadeh – Glean
Yeah, I think for us, as I mentioned at the top, all roads really lead back to PipeGen. It's the North Star metric for my entire team. Even Marketing Ops is constantly refreshing the dial on the Salesforce dashboard, even though they're not the ones with the exact ROI target. But the digital team in particular, who owns the website, really saw that pipeline was being left on the table because we couldn't rely on—
Sarah McConnell – Qualified
Mm-hmm.
Sarah McConnell – Qualified
What the—
Katy Abbaszadeh – Glean
—couldn't ask our human SDRs to be online 24/7. And so that team worked really closely with our Qualified CSM, Sean—shoutout—to put together a really compelling case to hire Piper. She is fully available on the website. We don't actually ask any of our human SDRs to be on that channel any longer. But I will say we don't plan on fully replacing our human SDRs. We really believe that there's an important place for them as well. But it was about pipeline. We didn't want to leave any dollar left behind.
Sarah McConnell – Qualified
I totally agree with you. And I feel like we went through the same sort of transformation here. I obviously get to be like on the forefront of our products and be our guinea pig, which is a blessing and a curse. We're the same way, and like asking our reps to be online all the time. Obviously one, can't—like they have to sleep, they have to eat. They cannot be online all the time. And even when they were, I just was just talking to someone else about tire deflection and like these, or sorry, tire kickers and like kind of deflection of people on the website is other, I'm like, am I really utilizing my SDR teams the right way when I'm seeing these conversations they're having with these people that aren't really that interested in like, yes, we can deflect them through the product. But I'm like, what if I—if I just have an AI agent do all of this, they can go focus on the things. And this is such a played out term, but like they can go focus on the things I want them to focus on and a lot more high-value things. Like cold outbounding is hard. It's a very hard thing to do. It's hard to do really well. We do have an outbound AI SDR that we use as well. But when it comes to like named accounts and really personalized experiences, I still very much believe in a human touch for that. So I'm like, if I can free up their time to do the good research and actually write really compelling messaging, is their time better spent doing that? So yeah, I agree with you on that. It's not about replacing them, but like, can we find areas for them that are gonna be more impactful for organization and for their career trajectory? Because responding to like mundane—
Katy Abbaszadeh – Glean
Thank you.
Sarah McConnell – Qualified
—answers on our website all day long, it's not really a great learning experience for them necessarily.
Katy Abbaszadeh – Glean
100%. And we still arm the human SDRs with their own AI for that personalization content generation. But I think another played-out term, human in the loop, is still really critical to success across the board.
Sarah McConnell – Qualified
Yeah.
Sarah McConnell – Qualified
And now beyond pipe generation, was there any unexpected things that bringing on an AI SDR agent helped unlock for your team? Is there anything you noticed that went beyond your expectations of pipeline generation and following up with leads?
Katy Abbaszadeh – Glean
Yeah, well, you actually just touched on it. And perhaps it shouldn't have been unexpected. But I do think the biggest eye-opener for me personally was the DQ rate. Piper does such a good job of weeding out prospects who just aren't a good fit, maybe yet or never will be. And so that saving for the human SDRs’ time was huge. So I think that was an eye-opener. And then just other metrics that we've been keeping a very close eye on with related to Piper is that time to first touch, you know, compared to the control of when the human SDRs were on when they could be, DQ rate, as I mentioned, and then meeting acceptance and show rate. Like, is there a difference between what Piper's able to book versus our SDR team? And then incremental meetings pipeline obviously is that outcome North Star that we're going for, and we've actually seen an uptick in that, I think due to the speed and efficiency, and she's always there.
Sarah McConnell – Qualified
That's incredible. You kind of already answered this, but I want to dig in a little bit deeper. As a marketing leader that's obviously taking AI very seriously within your organization, if you were talking to a peer and they're like, I just don't know where to start. I feel very overwhelmed. I think we've all felt that way at some point. What advice would you give to them on where to start? What is the best starting point to get over that hump of like, okay, I've run into too many walls with AI. I don't feel like this is worth my time. What advice would you give them?
Katy Abbaszadeh – Glean
This is a tough one because I think every organization is a little bit different, and depending on what your ultimate metric is, you want to make sure that your stake in the ground with AI has an impact on that. So I guess I would say, sort of back to the high-leverage use case, pick something that is a money-in, money-out workflow that you know like the back of your hand in the traditional way of marketing, and really try to map that out as specific jobs to be done and then rework that as an AI flow and do a comparison. Run a six-week A-B test and see what you get at the end. I think being clear about what you are measuring in that experiment is important, but then kind of let it do its thing and—what happens. I think that's the best way to really dip your toes in, and maybe in some cases traditional wins out because your underlying infrastructure isn't quite ready yet or, you know, I would say even if AI doesn't win the first battle, don't give up because there are always things that you can continue to refine and test.
Sarah McConnell – Qualified
And you said it earlier, I think it's okay to fail. I think that is something as marketers—don't, maybe it's just the circles that I run in—but marketers tend to be pretty type A for the most part. And I think failing is not something we love to do. I think it's a demand gen thing. I will say we tend to be cut from the same cloth. But failing is hard. Like I think if, as a demand gen person or someone that owns a pipeline number, you tend to be a little competitive. You've got big numbers that you're always looking to hit. You've got that drive in you, and failing, even if it is something that's new and different—
Katy Abbaszadeh – Glean
Guilty.
Katy Abbaszadeh – Glean
Yeah.
Sarah McConnell – Qualified
—can feel very frustrating. So I like that you said that early about be okay failing, but if you set the outcomes that you're looking for, it makes it a lot easier to fail fast and say like, okay, this isn't working, let's pivot, let's move on versus just spinning your wheels on something for a long time that ends up not playing out and/or giving up too fast. Like it might be something that you want to stick to. So I think your advice is great and also like fail quick and move on.
Katy Abbaszadeh – Glean
Yeah, I was just at an AI for Women meetup in Minneapolis last week, moderated a panel there. And I think that was my number one takeaway. It was really, we have to be able to embrace failure right now. And this wasn't marketing specific. It was kind of across the board—sales enablement, IT folks. But it's so new and changing so fast that if you expect it to be perfect and intuitive and something you can just roll out and walk away from—
Sarah McConnell – Qualified
Yeah.
Katy Abbaszadeh – Glean
—then you're definitely setting yourself up for failure. But if you take the bite-size failures along the way, like you said, fail fast and learn from it, then you're setting yourselves up—and your team—for success.
Sarah McConnell – Qualified
Absolutely. Now, as we think about AI in general, it's obviously made a massive impact on us as marketers and within our organizations. But I think it's also making a pretty big impact on the expectations of buyers. So do you have any advice for us as marketers? Like, how do you think AI is changing buyer expectations? And are we underestimating anything that we need to be cognizant of that AI is really changing for our buyers that we need to keep up with?
Katy Abbaszadeh – Glean
I'm gonna steal an answer from a CEO of a vendor I chatted with a few months ago, but it has just stuck with me. Eventually, you can imagine a world where we're marketing not to humans anymore, but to agents who bring back answers from queries on ChatGBT. Those agents are gonna go out to the websites, figure out how to compare different products and—
Sarah McConnell – Qualified
Mmm.
Katy Abbaszadeh – Glean
—serve up an answer where buyers expect a one-minute internet immediate answer—what's the best choice? And so I think we as marketers really need to be aware that some of our key metrics are going to shift. Should we care as much about website visitors in the future? Because it might just be an agent pinging your website or having it cached. So I don't know. It's a bit of a scary frontier because we're—
Sarah McConnell – Qualified
Yeah.
Katy Abbaszadeh – Glean
—we love our websites, don't we? But I do think we have to think about how our metrics shift, who we're marketing to shifts. So a lot to think about on that front.
Sarah McConnell – Qualified
I just had someone on the podcast last week and they kind of shared the same sentiment. And I think that's been like prolific around the marketing—just any peers I've talked to is like, yeah, website traffic is down. Like I think LLMs and the way that people are engaging with search is just changing so drastically, so quickly. But the continued sentiment has been the traffic that is coming from those, the AI sources, tends to be a much higher quality because they've already furthered themselves so much farther down the buying cycle, because of the way they're engaging with them. And at some point it's gonna be agents coming in, querying your website, and they're gonna come back with an answer as the buying cycles go faster. There are much higher conversion rates of those visitors coming from those sources converting into opportunities, and then once they're opportunities, they're closing quicker. As marketers, it will be so—it'll be a new challenge to convince our executive teams that like, yeah, it's okay that website traffic is dropping because the outputs and the metrics that we care about are still—as long as you're maintaining those—it's fine. But it is very scary to suddenly see this dramatic drop in traffic and sources. That's something I think we've all been struggling with just based on the conversations that I've been having.
Katy Abbaszadeh – Glean
Yeah, totally agree with that. And while the traffic from the LLMs is still small compared to everything else, to your point, the quality and the conversion and the speed is so much better. I think we just have to continue focusing on outcomes and that ultimate outcome that we're looking for versus the input metrics that we've typically kind of hung our hats on. So I think it makes it that much more important to drive that alignment with sales. What is the definition of pipeline? At what point, at what stage in the cycle are we measuring it? What coverage do we need? And making sure that we as marketers are speaking that same language so that we're not over here saying, the traffic, but the MQL.
Sarah McConnell – Qualified
Hahaha. Now, on that same note, like we just talked about, obviously search and web traffic has been a huge change, but is there another big—like, what would you say is gonna be the single biggest change to B2B marketing in the next 12 months, or is it going to be search in the way that people are interacting with our website?
Katy Abbaszadeh – Glean
Yeah, predicting the future. Yes, here we go. You know, I think this is less in my wheelhouse from a demand perspective, but as a buyer myself, I'm feeling like this shift is something we can't ignore. Pricing structure and pricing expectations. People do not want to get locked into—
Sarah McConnell – Qualified
I know, I'm like, hey, can you get out your crystal ball and just like tell me what the future holds, please?
Sarah McConnell – Qualified
Mm-hmm.
Sarah McConnell – Qualified
Yeah.
Katy Abbaszadeh – Glean
—annual contracts because things are changing so fast and you want to be able to test. So I think it's maybe a little bit easier to price that way in a consumer market, but on the B2B side, we have to be thinking about how we shift to that expectation so that people can be confident, you know, testing and using a product without getting locked in for one- to two-year contracts.
Sarah McConnell – Qualified
Yeah, I think that's a really good one. And earlier we've kind of talked—I think we've touched a lot throughout this conversation on use cases for AI agents. Like, I think you gave some really great examples that you're seeing within your own marketing organization. But one of the things I'm so curious about is what do you hope to see more of soon? Like, are there any areas within your marketing team that you're like, man, I really wish there was some sort of agentic solution that could help me scale this that you're just not seeing as much of in the market right now?
Katy Abbaszadeh – Glean
Yeah, this one's maybe less marketing focused, but I really am liking the agents that are more focused on coaching. So, executive coaching—I'm, you know, moving up in my career. How can I make sure I'm speaking to the right audience and the right tone and saying the right things? For our sales team, first pitch coaching. So I think that will probably become more formalized. It's a really interesting use case. One that's not necessarily a use case, but I do think we need to get there as an industry is open metrics standards, so that we can actually compare agent performance apples to apples. Obviously, each company is going to have a tech stack, different way of building agents. But if there's a way that we can—
Sarah McConnell – Qualified
Hmm.
Katy Abbaszadeh – Glean
—as marketers, always asking, OK, what's the industry average for CAC and web visits and all of that? We need a similar way to benchmark how we're doing in AI, and I don't think we're quite there yet.
Sarah McConnell – Qualified
That's probably my favorite answer I've heard today. And as you said, industry benchmarks, I like had a shiver run through me because I think in demand gen, the number of times we're like, well, what's the industry after—I'm like, I don't, I don't know. Like it's so different across organizations. Like in every company I've worked at, it's been so different. I tried to crowdsource recently, like as website traffic has been dropping, I'm like, where are people seeing—like what's your percentage breakdowns of web traffic? And it was—when I say all over the board is an understatement. So being able to have more standardized averages, especially as we think about AI, like, I like jump for joy. I was like, yes, can someone please make this if you're listening, because we just don't have that insight. And I feel like we're always flying a little bit blind.
Katy Abbaszadeh – Glean
Yeah, well hopefully the LLMs can already help because they've got access to all of the information. So you can get more specific on what do we even mean by industry when we say industry average. So I think that's promising, but we gotta do some work to make sure it's there for everyone.
Sarah McConnell – Qualified
Totally. Okay, my last question before we get into my favorite section, which is the lightning round. I'm curious if there is something that you think most B2B orgs will still get wrong about AI a year from now. So before I ask you—you know, get out your crystal ball, pull it back out—what do you think we're getting wrong now that you don't think is gonna get better? Or that, I guess I want to phrase that is, what do you think we should be focusing on because you're worried it's a problem or that we're still getting wrong?
Katy Abbaszadeh – Glean
Yeah, I'll do a callback to one that I kind of mentioned earlier, but I think it's about focusing too much or celebrating too much the demoable moments, the like little delightful assistance that like, hey, look how cute this is, and ignoring what that boring backbone takes to make it work more in a complex manner. So the data governance—at that breakfast I was at last week, there was a ton of talk about changing the terminology of data governance to AI governance. And how do we—you know, another phrase we like to use, garbage in, garbage out—how do we make sure that the AI is set up for success? And building that foundation is hard. And there's a lot of question on who owns building that foundation. You know, when you're in marketing and—
Sarah McConnell – Qualified
Mm-hmm.
Katy Abbaszadeh – Glean
—you're partnering with rev ops or IT, you know, who really owns that and how do you make sure that's set up appropriately? I think that'll be the tough thing for people to get right. So I'm interested to see how we solve that problem collectively.
Sarah McConnell – Qualified
Yeah, that is a great answer. Okay, I love to wrap up these episodes with our lightning round. So it's quick questions with quick answers. So to get started, other than ChatGPT, what was one of the first AI tools that you experimented with?
Katy Abbaszadeh – Glean
This is going to sound really lame. I promise I'm not going to sell Glean. But it was Glean because I was just coming back from maternity leave right after ChatGBT exploded onto the scene. And our engineering team had just done a hackathon. And three weeks after the hackathon, we launched Glean Chat. And so I really played with ChatGBT through Glean before I even played directly with ChatGBT.
Sarah McConnell – Qualified
That is such a—what a wild thing to come back to maternity leave from is like, I get back from maternity leave in August and I was like, what is vibe coding? Like, can someone—that's been four months and I feel like suddenly it was a completely different world. And that was the least dirty day. I feel like you came back to like a whole—literally a whole new world. Okay, what do you think is the most overrated buzzword in martech right now?
Katy Abbaszadeh – Glean
Right? Exact same experience. Yes, it was insane. I'm gonna go with personalization at scale. And I think this one is one that we like to use. It sounds great. We all talk about ABM. But without that governance piece and a measurement on outcome, it's just louder spray and pray. So.
Sarah McConnell – Qualified
It's so true. Okay, who is a marketer that you think is ahead of the curve on AI that listeners should go follow on LinkedIn right now?
Katy Abbaszadeh – Glean
OK, I will give you a name, but I first want to start with a shoutout to another podcast that I recently found. So on top of this podcast, which obviously you're already following because you're listening to this, Humans of MarTech has been a really great one that I've recently discovered. So shoutout to them, curating a ton of really interesting content from marketing leaders. And then, for an individual person, I have to give a shoutout to our head of digital marketing, Daniel Henderson. He has been partnering closely with Kevin Indig, who is somewhat of an AISCO influencer out there. And he's been involved in a lot of our AISCO projects. He's got the scrappy, rapid testing approach that has really helped us kind of frame our thinking and where we want to place bets. So Kevin Indig is a good one to check out.
Sarah McConnell – Qualified
Amazing. Okay, last question. If you could automate one part of your life with AI that's outside of work, what would it be?
Katy Abbaszadeh – Glean
I would absolutely love an AI stylist or personal shopper. And I'm sure there's a version of this that probably already exists, or I could go train my own custom GPT, but who has the time for that?
Sarah McConnell – Qualified
Absolutely. So I agree with that. Okay, Katy, I appreciate you joining us so much. Thank you for joining The Agentic Marketer podcast. It was great having you on and having you your insights. So thank you so much.
Katy Abbaszadeh – Glean
Thank you, Sarah.
Stay up to date with weekly drops of fresh B2B marketing and sales content.
Discover how Katy Abbaszadeh at Glean uses agentic AI to scale outcomes, boost pipeline ROI, and transform traditional marketing.
This episode features an interview with Katy Abbaszadeh, Head of Pipeline Marketing at Glean, a leading work AI platform that connects company apps and data to help teams find information, collaborate, and make better decisions faster.
Katy shares how Glean is redefining pipeline generation with agentic AI, using autonomous agents to scale outcomes instead of just activities. She discusses how her team moved from “cute AI experiments” to leading with AI-first strategies that drive measurable impact, and what it takes to build a culture that embraces automation, experimentation, and continuous learning.
Key Takeaways:
TRANSCRIPT
Sarah McConnell – Qualified
Okay, Katy, thank you so much for joining us on The Agentic Marketer, where we're taking a peek into today's leaders’ tech stacks and AI strategies to learn how you're using AI agents to hit your pipe targets. But before we dive in, Katy, I'd love for you just to tell us a little bit about yourself and the work that you're doing over at Glean.
Katy Abbaszadeh – Glean
Sure, I'd be happy to. And first off, thanks for having me. Excited to be here. So I'm Katy Abbaszadeh. I'm responsible for what I refer to as all things pipeline marketing here at Glean. That includes demand gen, events, operations, really anything that's turning leads into meetings into pipeline. Yeah, go ahead.
Sarah McConnell – Qualified
Oh no, I love that. I'm thinking about my role, and I'm like, yeah, that is exactly what it is. Like all's I care about is pipeline.
Katy Abbaszadeh – Glean
Yep, that's the ultimate metric for us. And really the charter is high-quality pipeline at industry-leading ROI.
Sarah McConnell – Qualified
That's amazing. And now, obviously the whole show we're talking about agentic marketing. So obviously you're into ManGen, you're thinking about pipeline all the time, but I'm curious to kick things off. Like how do you define agentic marketing? Like what does that term mean to you and the team over at Glean?
Katy Abbaszadeh – Glean
Yeah, I think we're all at different stages in our AI and agentic roadmap, so to speak. So I'm so glad you have this forum for folks to share where they're at and what they're doing with it. I think for us, the future vision and where we're starting to get a foothold is that agentic marketing really means that we're orchestrating autonomous and semi-autonomous agents to do work for us. So it's about grounding the agents in your own data, but then making sure it's measured on things like revenue impact. We as marketers really care about the outcomes, not just the inputs. So making sure that the agents that you have working on your behalf have an outcome that they're working towards. So I think about it as not just implementing a set of new age tools, but really a new operating model for how we plan work, execute work, staff work.
Sarah McConnell – Qualified
Yeah, that's such a good answer. As I think about my next question is like, how is it different than traditional marketing? And at the, guess the base of it, seems pretty obvious, like obviously talking to AI and agents and that's very different. But do you feel like agentic marketing is vastly different than traditional marketing, or is it at its base the same thing we've always been doing just with like a different twist to it? But how do you think about those two things differently?
Katy Abbaszadeh – Glean
It's a good question. I think more recently I've come to believe that it is a bigger shift from traditional. When it was first coming out, I was like, that's cute that AI can do that. But now I'm more on the lines of thinking, this is about moving from in traditional marketing where we were scaling humans with tools. Now we're taking agents that can scale outcomes with systems. So thinking about moving from activities to that outcome base. You'll hear me talk about that a lot. It's focusing on what the end result is. Channels to agentic chains, different agents that can do different things, and fully autonomously is the ultimate vision. I'll admit we're not quite there yet, but there's a lot of promise there. And then from individual ops, QAing on a human level, to governance, systems of governance that can kind of be the guardrails to what your agents are doing.
Sarah McConnell – Qualified
Yeah, and I really like that answer because I feel like in 2023, when it was first like generative AI, that was like, oh, this is cute. Like it helps me. It's a little, we used to call the co-pilot. So you're riding shotgun and helping me make my job a little bit easier, but it just didn't feel like this massive shift. It was just like a slight change. But then now here in 2025, I agree. I'm like, oh no, this is vastly different. And we are being asked as marketers to have a completely different mindset and framework because we now do have these autonomous agents that we need to orchestrate and we need to have them do full roles that we are holding them accountable to an output, and that does feel vastly different than what we used to do. So I think your answer is perfect in that sense.
Katy Abbaszadeh – Glean
Yeah, 100 percent. I would say even—not to cut you off—but as we're heading into next year planning, you know, I'm starting to ask my managers, when you're thinking about headcount for next year, how does AI play into that? Not to the full extent of like we're freezing hiring until you prove that AI can't do this, but really thinking about AI as part of your team. And I haven't created a prototype prototype yet, but, you know, what does the new org chart look like, and how do you highlight the agents that are doing work?
Sarah McConnell – Qualified
That's a good segue to my next question, which I'm so curious. Making that leap from experimentation with AI to truly leading with it, I think in the way you just spoke about it, you are truly thinking about this as, 2026 planning, how is AI scaling your team? How are you thinking about bringing them on to help already existing org structures? So that is a big jump. So what did it take over at Glean for you to feel like you're making that move from the like, this is cute, this co-pilot is helping, to like, okay, no. This is a part of our 2026 planning. We need to lead with agentic AI and agents.
Katy Abbaszadeh – Glean
Yeah, it's a good question. And frankly, I don't think it's going to just be one single leap. I think it's actually baby steps to get to that ultimate leap. Any new skill, there's a learning curve. We're all adopting this new thing in real time. It's changing, feels like daily, weekly, new stuff popping on the scene. So I definitely think we have to give ourselves grace. And I try to think about leading with vulnerability. Not everything we test is going to come out the way we want it to. Not everything is going to be a smashing success. So how do we think about, as leaders, rolling up our sleeves, getting in the mix, being on the vendor calls, you know, workflow write-outs of jobs to be done, where can AI fit in? I think that's a really important piece of it. And I meant to introduce Glean lightly at the start, just because I think it helps to have some context of what Glean is. We're a work AI platform. So I have the benefit of working for an organization that is trying to transform work with AI across the board. So we're a horizontal platform that connects all of your apps and data at work in a safe way. So I have a bit of a cheat code in that my product that I'm marketing actually helps me be an AI-first marketer. But back to your question about what does that leap—I think the practical answer I'll give is I really recommend starting with one high-leverage workflow that you can really easily map out from what you're doing today and then kind of build in what are those things that can be replaced by AI. So an example is an ABM activation. Obviously, we hear a lot about content generation for AI being a major use case, personalization. So—
Sarah McConnell – Qualified
You're good.
Katy Abbaszadeh – Glean
—what is one ABM activation you're already doing that you can now kind of shift and think about it in an AI framework.
Sarah McConnell – Qualified
Yeah, I think that you talked about Glean being an AI-first company and giving you sort of that AI-first mindset, and obviously working at Qualified, I feel the same way. We have an AI-native product. We have our AI SDR, Piper, which we'll talk about, but it is such a unique position to be in because it does give you sort of the permission to really lead with that AI-first mentality, which I don't think is necessarily the case at every single organization. So I think to your point of starting there and it's starting with that experimentation, but picking like one thing, but being very lucky to work for organizations that do—
Katy Abbaszadeh – Glean
See you on the next video.
Sarah McConnell – Qualified
I'm assuming like your leadership team is asking you, like, where is AI showing up in your day-to-day because you're an AI product? So you're saying help to that accountability—
Katy Abbaszadeh – Glean
Yep. Yeah, absolutely. It takes forcing yourself to think about it every day. And I think at first, like we talked about it, it's so cute. It's fun. It's an experiment, right? But now, still maintaining time and space for the team to think about it, get creative with it, test and try and fail. I think the failures are really important in these baby steps, ultimate leap. So constantly trying to keep it top of mind.
Sarah McConnell – Qualified
Absolutely. Now, I'm curious, you said you oversee DemandGen, you oversee events and ops. Where is agentic AI showing up the most visibly on your team, whether that be from a DemandGen perspective or across your largest, like the marketing org as a whole? I think listeners are always so curious of, okay, agentic AI, agentic marketing sounds really cool, but I don't know where to put this into play. So where is it showing up for you on your Glean team most prominently?
Katy Abbaszadeh – Glean
Yeah, I think generally it's in those outcomes we talked about earlier, what we're kind of referring to as revenue moments. So you'll hear it come up when we get to the Piper part of the discussion, but that inbound speed to lead piece is really important to us. Outbound sequencing, some of those ABM moments where we can help the human SDRs that we do have be more personalized and proactive. Content generation and personalization we talked about already. Pre-meeting research for our reps that we as marketers can kind of curate for them. And then that, back to the inbound speed to lead, that 24/7 coverage that we just don't have to worry somebody's not going to get that custom touch that we want them to have. Obviously, productivity is a huge win across the board. So just planning time for campaign output.
Sarah McConnell – Qualified
Them.
Katy Abbaszadeh – Glean
You know, we've seen some of our planning cycles go from something like 15 days to a day because we can feed an agent the ideas, brainstorm with it, have a plan right then and there. And then more specifically within my team, I would say our digital marketing team is really leaning in heavily. So from a content generation perspective, we've got our SEO strategy really humming with AI workflows. Obviously, Piper is a big piece of that, our AI SDR on the website. And then even getting into some vibe coding on new website modules. So if you check out our homepage, you can see some that were vibe coded and then kind of built into the glean.com homepage. So some fun stuff there. And then the ABM stuff is a big play as well.
Sarah McConnell – Qualified
That's amazing. It's so cool to hear how many different aspects of your organization are adopting this, because I feel like getting it into one part of the organization always feels like that first step, and that can be not easy, but is easier to do, but getting it to be sort of an org-wide mindset, having been on the forefront of that, is much, much harder. And then my last question of this section of our conversation today, I'm very curious, do you have a hot take about AI and marketing that most people would not agree with you on?
Katy Abbaszadeh – Glean
Oof, hot takes. Love a hot take. I have to go back to the cutesy theme of how it burst onto the scene. I think right now most teams are overestimating the novelty and underestimating the operations portion of it, the backend portion of it. Agents are ready, and I'm not saying that Glean has it down perfectly just yet. They're still learning. But agents are ready, and it's really about—
Sarah McConnell – Qualified
Hahaha!
Sarah McConnell – Qualified
Them.
Katy Abbaszadeh – Glean
—org design, underlying data governance, how you're measuring things that become the bottlenecks to success. So I think the winners of this wave, so to speak, will really treat this as a go-to-market transformation and not just spot tools that you're rolling out here and there, but really rethinking how you plan, staff, and actually execute the work.
Sarah McConnell – Qualified
That's such a great answer because I do think it is a hot take because I do think we're hearing more in the market right now of people being like, well, maybe it's not as impactful as we thought it was going to be. And I worry that it's going to make marketers kind of rest on their laurels a little bit and be like, well, you know, if other people aren't finding success with it or I'm saying, you know, that's not impacting our bottom line, I don't want to invest the time and the effort to get up to speed with agentic AI and how I'm going to bring it onto my team. And I worry that with those two shifting narratives of like, no, it is really important versus like, OK, maybe there's not as much impact of it is that too many marketers will wait and say like, it isn't going to be that impactful. And then when it is, we'll feel left behind. So I love that your take is like, no, it's here. It's ready. It's kind of on us to build that infrastructure into our team. And I do, I agree with you. I think the teams that are embracing it early and are building in that infrastructure are the ones that are going to win and be way ahead. All of we're like, no, this is here to stay. This is going to be a vital part of our organizations. Encouraging other marketers to stay up to date with it and stay up to speed I think is so, so important.
Katy Abbaszadeh – Glean
Yeah. And I just want to say I understand the feeling of maybe burnout or, I tried this thing that I thought was going to be so great and I hit a wall and why would I try it again? I think even just having the ideas for, you know, certain workflows or what will be life changing when AI meets you there, having that scaffolding ready to go when AI catches up, you're that much further ahead. So. Keep it up.
Sarah McConnell – Qualified
Totally. Okay, so shifting more gears into what you guys are doing over at Glean, you guys have hired Piper, the AI SDR from Qualified, which is amazing. We're so happy to have you guys as customers. You've renamed her Taylor. What was it about having an AI SDR agent that made you and the team lean in? Like, what was it about that particular use case that the Glean team was like, okay, this is something I think we can start with or we can start utilizing within our organization?
Katy Abbaszadeh – Glean
Yeah, I think for us, as I mentioned at the top, all roads really lead back to PipeGen. It's the North Star metric for my entire team. Even Marketing Ops is constantly refreshing the dial on the Salesforce dashboard, even though they're not the ones with the exact ROI target. But the digital team in particular, who owns the website, really saw that pipeline was being left on the table because we couldn't rely on—
Sarah McConnell – Qualified
Mm-hmm.
Sarah McConnell – Qualified
What the—
Katy Abbaszadeh – Glean
—couldn't ask our human SDRs to be online 24/7. And so that team worked really closely with our Qualified CSM, Sean—shoutout—to put together a really compelling case to hire Piper. She is fully available on the website. We don't actually ask any of our human SDRs to be on that channel any longer. But I will say we don't plan on fully replacing our human SDRs. We really believe that there's an important place for them as well. But it was about pipeline. We didn't want to leave any dollar left behind.
Sarah McConnell – Qualified
I totally agree with you. And I feel like we went through the same sort of transformation here. I obviously get to be like on the forefront of our products and be our guinea pig, which is a blessing and a curse. We're the same way, and like asking our reps to be online all the time. Obviously one, can't—like they have to sleep, they have to eat. They cannot be online all the time. And even when they were, I just was just talking to someone else about tire deflection and like these, or sorry, tire kickers and like kind of deflection of people on the website is other, I'm like, am I really utilizing my SDR teams the right way when I'm seeing these conversations they're having with these people that aren't really that interested in like, yes, we can deflect them through the product. But I'm like, what if I—if I just have an AI agent do all of this, they can go focus on the things. And this is such a played out term, but like they can go focus on the things I want them to focus on and a lot more high-value things. Like cold outbounding is hard. It's a very hard thing to do. It's hard to do really well. We do have an outbound AI SDR that we use as well. But when it comes to like named accounts and really personalized experiences, I still very much believe in a human touch for that. So I'm like, if I can free up their time to do the good research and actually write really compelling messaging, is their time better spent doing that? So yeah, I agree with you on that. It's not about replacing them, but like, can we find areas for them that are gonna be more impactful for organization and for their career trajectory? Because responding to like mundane—
Katy Abbaszadeh – Glean
Thank you.
Sarah McConnell – Qualified
—answers on our website all day long, it's not really a great learning experience for them necessarily.
Katy Abbaszadeh – Glean
100%. And we still arm the human SDRs with their own AI for that personalization content generation. But I think another played-out term, human in the loop, is still really critical to success across the board.
Sarah McConnell – Qualified
Yeah.
Sarah McConnell – Qualified
And now beyond pipe generation, was there any unexpected things that bringing on an AI SDR agent helped unlock for your team? Is there anything you noticed that went beyond your expectations of pipeline generation and following up with leads?
Katy Abbaszadeh – Glean
Yeah, well, you actually just touched on it. And perhaps it shouldn't have been unexpected. But I do think the biggest eye-opener for me personally was the DQ rate. Piper does such a good job of weeding out prospects who just aren't a good fit, maybe yet or never will be. And so that saving for the human SDRs’ time was huge. So I think that was an eye-opener. And then just other metrics that we've been keeping a very close eye on with related to Piper is that time to first touch, you know, compared to the control of when the human SDRs were on when they could be, DQ rate, as I mentioned, and then meeting acceptance and show rate. Like, is there a difference between what Piper's able to book versus our SDR team? And then incremental meetings pipeline obviously is that outcome North Star that we're going for, and we've actually seen an uptick in that, I think due to the speed and efficiency, and she's always there.
Sarah McConnell – Qualified
That's incredible. You kind of already answered this, but I want to dig in a little bit deeper. As a marketing leader that's obviously taking AI very seriously within your organization, if you were talking to a peer and they're like, I just don't know where to start. I feel very overwhelmed. I think we've all felt that way at some point. What advice would you give to them on where to start? What is the best starting point to get over that hump of like, okay, I've run into too many walls with AI. I don't feel like this is worth my time. What advice would you give them?
Katy Abbaszadeh – Glean
This is a tough one because I think every organization is a little bit different, and depending on what your ultimate metric is, you want to make sure that your stake in the ground with AI has an impact on that. So I guess I would say, sort of back to the high-leverage use case, pick something that is a money-in, money-out workflow that you know like the back of your hand in the traditional way of marketing, and really try to map that out as specific jobs to be done and then rework that as an AI flow and do a comparison. Run a six-week A-B test and see what you get at the end. I think being clear about what you are measuring in that experiment is important, but then kind of let it do its thing and—what happens. I think that's the best way to really dip your toes in, and maybe in some cases traditional wins out because your underlying infrastructure isn't quite ready yet or, you know, I would say even if AI doesn't win the first battle, don't give up because there are always things that you can continue to refine and test.
Sarah McConnell – Qualified
And you said it earlier, I think it's okay to fail. I think that is something as marketers—don't, maybe it's just the circles that I run in—but marketers tend to be pretty type A for the most part. And I think failing is not something we love to do. I think it's a demand gen thing. I will say we tend to be cut from the same cloth. But failing is hard. Like I think if, as a demand gen person or someone that owns a pipeline number, you tend to be a little competitive. You've got big numbers that you're always looking to hit. You've got that drive in you, and failing, even if it is something that's new and different—
Katy Abbaszadeh – Glean
Guilty.
Katy Abbaszadeh – Glean
Yeah.
Sarah McConnell – Qualified
—can feel very frustrating. So I like that you said that early about be okay failing, but if you set the outcomes that you're looking for, it makes it a lot easier to fail fast and say like, okay, this isn't working, let's pivot, let's move on versus just spinning your wheels on something for a long time that ends up not playing out and/or giving up too fast. Like it might be something that you want to stick to. So I think your advice is great and also like fail quick and move on.
Katy Abbaszadeh – Glean
Yeah, I was just at an AI for Women meetup in Minneapolis last week, moderated a panel there. And I think that was my number one takeaway. It was really, we have to be able to embrace failure right now. And this wasn't marketing specific. It was kind of across the board—sales enablement, IT folks. But it's so new and changing so fast that if you expect it to be perfect and intuitive and something you can just roll out and walk away from—
Sarah McConnell – Qualified
Yeah.
Katy Abbaszadeh – Glean
—then you're definitely setting yourself up for failure. But if you take the bite-size failures along the way, like you said, fail fast and learn from it, then you're setting yourselves up—and your team—for success.
Sarah McConnell – Qualified
Absolutely. Now, as we think about AI in general, it's obviously made a massive impact on us as marketers and within our organizations. But I think it's also making a pretty big impact on the expectations of buyers. So do you have any advice for us as marketers? Like, how do you think AI is changing buyer expectations? And are we underestimating anything that we need to be cognizant of that AI is really changing for our buyers that we need to keep up with?
Katy Abbaszadeh – Glean
I'm gonna steal an answer from a CEO of a vendor I chatted with a few months ago, but it has just stuck with me. Eventually, you can imagine a world where we're marketing not to humans anymore, but to agents who bring back answers from queries on ChatGBT. Those agents are gonna go out to the websites, figure out how to compare different products and—
Sarah McConnell – Qualified
Mmm.
Katy Abbaszadeh – Glean
—serve up an answer where buyers expect a one-minute internet immediate answer—what's the best choice? And so I think we as marketers really need to be aware that some of our key metrics are going to shift. Should we care as much about website visitors in the future? Because it might just be an agent pinging your website or having it cached. So I don't know. It's a bit of a scary frontier because we're—
Sarah McConnell – Qualified
Yeah.
Katy Abbaszadeh – Glean
—we love our websites, don't we? But I do think we have to think about how our metrics shift, who we're marketing to shifts. So a lot to think about on that front.
Sarah McConnell – Qualified
I just had someone on the podcast last week and they kind of shared the same sentiment. And I think that's been like prolific around the marketing—just any peers I've talked to is like, yeah, website traffic is down. Like I think LLMs and the way that people are engaging with search is just changing so drastically, so quickly. But the continued sentiment has been the traffic that is coming from those, the AI sources, tends to be a much higher quality because they've already furthered themselves so much farther down the buying cycle, because of the way they're engaging with them. And at some point it's gonna be agents coming in, querying your website, and they're gonna come back with an answer as the buying cycles go faster. There are much higher conversion rates of those visitors coming from those sources converting into opportunities, and then once they're opportunities, they're closing quicker. As marketers, it will be so—it'll be a new challenge to convince our executive teams that like, yeah, it's okay that website traffic is dropping because the outputs and the metrics that we care about are still—as long as you're maintaining those—it's fine. But it is very scary to suddenly see this dramatic drop in traffic and sources. That's something I think we've all been struggling with just based on the conversations that I've been having.
Katy Abbaszadeh – Glean
Yeah, totally agree with that. And while the traffic from the LLMs is still small compared to everything else, to your point, the quality and the conversion and the speed is so much better. I think we just have to continue focusing on outcomes and that ultimate outcome that we're looking for versus the input metrics that we've typically kind of hung our hats on. So I think it makes it that much more important to drive that alignment with sales. What is the definition of pipeline? At what point, at what stage in the cycle are we measuring it? What coverage do we need? And making sure that we as marketers are speaking that same language so that we're not over here saying, the traffic, but the MQL.
Sarah McConnell – Qualified
Hahaha. Now, on that same note, like we just talked about, obviously search and web traffic has been a huge change, but is there another big—like, what would you say is gonna be the single biggest change to B2B marketing in the next 12 months, or is it going to be search in the way that people are interacting with our website?
Katy Abbaszadeh – Glean
Yeah, predicting the future. Yes, here we go. You know, I think this is less in my wheelhouse from a demand perspective, but as a buyer myself, I'm feeling like this shift is something we can't ignore. Pricing structure and pricing expectations. People do not want to get locked into—
Sarah McConnell – Qualified
I know, I'm like, hey, can you get out your crystal ball and just like tell me what the future holds, please?
Sarah McConnell – Qualified
Mm-hmm.
Sarah McConnell – Qualified
Yeah.
Katy Abbaszadeh – Glean
—annual contracts because things are changing so fast and you want to be able to test. So I think it's maybe a little bit easier to price that way in a consumer market, but on the B2B side, we have to be thinking about how we shift to that expectation so that people can be confident, you know, testing and using a product without getting locked in for one- to two-year contracts.
Sarah McConnell – Qualified
Yeah, I think that's a really good one. And earlier we've kind of talked—I think we've touched a lot throughout this conversation on use cases for AI agents. Like, I think you gave some really great examples that you're seeing within your own marketing organization. But one of the things I'm so curious about is what do you hope to see more of soon? Like, are there any areas within your marketing team that you're like, man, I really wish there was some sort of agentic solution that could help me scale this that you're just not seeing as much of in the market right now?
Katy Abbaszadeh – Glean
Yeah, this one's maybe less marketing focused, but I really am liking the agents that are more focused on coaching. So, executive coaching—I'm, you know, moving up in my career. How can I make sure I'm speaking to the right audience and the right tone and saying the right things? For our sales team, first pitch coaching. So I think that will probably become more formalized. It's a really interesting use case. One that's not necessarily a use case, but I do think we need to get there as an industry is open metrics standards, so that we can actually compare agent performance apples to apples. Obviously, each company is going to have a tech stack, different way of building agents. But if there's a way that we can—
Sarah McConnell – Qualified
Hmm.
Katy Abbaszadeh – Glean
—as marketers, always asking, OK, what's the industry average for CAC and web visits and all of that? We need a similar way to benchmark how we're doing in AI, and I don't think we're quite there yet.
Sarah McConnell – Qualified
That's probably my favorite answer I've heard today. And as you said, industry benchmarks, I like had a shiver run through me because I think in demand gen, the number of times we're like, well, what's the industry after—I'm like, I don't, I don't know. Like it's so different across organizations. Like in every company I've worked at, it's been so different. I tried to crowdsource recently, like as website traffic has been dropping, I'm like, where are people seeing—like what's your percentage breakdowns of web traffic? And it was—when I say all over the board is an understatement. So being able to have more standardized averages, especially as we think about AI, like, I like jump for joy. I was like, yes, can someone please make this if you're listening, because we just don't have that insight. And I feel like we're always flying a little bit blind.
Katy Abbaszadeh – Glean
Yeah, well hopefully the LLMs can already help because they've got access to all of the information. So you can get more specific on what do we even mean by industry when we say industry average. So I think that's promising, but we gotta do some work to make sure it's there for everyone.
Sarah McConnell – Qualified
Totally. Okay, my last question before we get into my favorite section, which is the lightning round. I'm curious if there is something that you think most B2B orgs will still get wrong about AI a year from now. So before I ask you—you know, get out your crystal ball, pull it back out—what do you think we're getting wrong now that you don't think is gonna get better? Or that, I guess I want to phrase that is, what do you think we should be focusing on because you're worried it's a problem or that we're still getting wrong?
Katy Abbaszadeh – Glean
Yeah, I'll do a callback to one that I kind of mentioned earlier, but I think it's about focusing too much or celebrating too much the demoable moments, the like little delightful assistance that like, hey, look how cute this is, and ignoring what that boring backbone takes to make it work more in a complex manner. So the data governance—at that breakfast I was at last week, there was a ton of talk about changing the terminology of data governance to AI governance. And how do we—you know, another phrase we like to use, garbage in, garbage out—how do we make sure that the AI is set up for success? And building that foundation is hard. And there's a lot of question on who owns building that foundation. You know, when you're in marketing and—
Sarah McConnell – Qualified
Mm-hmm.
Katy Abbaszadeh – Glean
—you're partnering with rev ops or IT, you know, who really owns that and how do you make sure that's set up appropriately? I think that'll be the tough thing for people to get right. So I'm interested to see how we solve that problem collectively.
Sarah McConnell – Qualified
Yeah, that is a great answer. Okay, I love to wrap up these episodes with our lightning round. So it's quick questions with quick answers. So to get started, other than ChatGPT, what was one of the first AI tools that you experimented with?
Katy Abbaszadeh – Glean
This is going to sound really lame. I promise I'm not going to sell Glean. But it was Glean because I was just coming back from maternity leave right after ChatGBT exploded onto the scene. And our engineering team had just done a hackathon. And three weeks after the hackathon, we launched Glean Chat. And so I really played with ChatGBT through Glean before I even played directly with ChatGBT.
Sarah McConnell – Qualified
That is such a—what a wild thing to come back to maternity leave from is like, I get back from maternity leave in August and I was like, what is vibe coding? Like, can someone—that's been four months and I feel like suddenly it was a completely different world. And that was the least dirty day. I feel like you came back to like a whole—literally a whole new world. Okay, what do you think is the most overrated buzzword in martech right now?
Katy Abbaszadeh – Glean
Right? Exact same experience. Yes, it was insane. I'm gonna go with personalization at scale. And I think this one is one that we like to use. It sounds great. We all talk about ABM. But without that governance piece and a measurement on outcome, it's just louder spray and pray. So.
Sarah McConnell – Qualified
It's so true. Okay, who is a marketer that you think is ahead of the curve on AI that listeners should go follow on LinkedIn right now?
Katy Abbaszadeh – Glean
OK, I will give you a name, but I first want to start with a shoutout to another podcast that I recently found. So on top of this podcast, which obviously you're already following because you're listening to this, Humans of MarTech has been a really great one that I've recently discovered. So shoutout to them, curating a ton of really interesting content from marketing leaders. And then, for an individual person, I have to give a shoutout to our head of digital marketing, Daniel Henderson. He has been partnering closely with Kevin Indig, who is somewhat of an AISCO influencer out there. And he's been involved in a lot of our AISCO projects. He's got the scrappy, rapid testing approach that has really helped us kind of frame our thinking and where we want to place bets. So Kevin Indig is a good one to check out.
Sarah McConnell – Qualified
Amazing. Okay, last question. If you could automate one part of your life with AI that's outside of work, what would it be?
Katy Abbaszadeh – Glean
I would absolutely love an AI stylist or personal shopper. And I'm sure there's a version of this that probably already exists, or I could go train my own custom GPT, but who has the time for that?
Sarah McConnell – Qualified
Absolutely. So I agree with that. Okay, Katy, I appreciate you joining us so much. Thank you for joining The Agentic Marketer podcast. It was great having you on and having you your insights. So thank you so much.
Katy Abbaszadeh – Glean
Thank you, Sarah.
Stay up to date with weekly drops of fresh B2B marketing and sales content.
Discover how Katy Abbaszadeh at Glean uses agentic AI to scale outcomes, boost pipeline ROI, and transform traditional marketing.
This episode features an interview with Katy Abbaszadeh, Head of Pipeline Marketing at Glean, a leading work AI platform that connects company apps and data to help teams find information, collaborate, and make better decisions faster.
Katy shares how Glean is redefining pipeline generation with agentic AI, using autonomous agents to scale outcomes instead of just activities. She discusses how her team moved from “cute AI experiments” to leading with AI-first strategies that drive measurable impact, and what it takes to build a culture that embraces automation, experimentation, and continuous learning.
Key Takeaways:
TRANSCRIPT
Sarah McConnell – Qualified
Okay, Katy, thank you so much for joining us on The Agentic Marketer, where we're taking a peek into today's leaders’ tech stacks and AI strategies to learn how you're using AI agents to hit your pipe targets. But before we dive in, Katy, I'd love for you just to tell us a little bit about yourself and the work that you're doing over at Glean.
Katy Abbaszadeh – Glean
Sure, I'd be happy to. And first off, thanks for having me. Excited to be here. So I'm Katy Abbaszadeh. I'm responsible for what I refer to as all things pipeline marketing here at Glean. That includes demand gen, events, operations, really anything that's turning leads into meetings into pipeline. Yeah, go ahead.
Sarah McConnell – Qualified
Oh no, I love that. I'm thinking about my role, and I'm like, yeah, that is exactly what it is. Like all's I care about is pipeline.
Katy Abbaszadeh – Glean
Yep, that's the ultimate metric for us. And really the charter is high-quality pipeline at industry-leading ROI.
Sarah McConnell – Qualified
That's amazing. And now, obviously the whole show we're talking about agentic marketing. So obviously you're into ManGen, you're thinking about pipeline all the time, but I'm curious to kick things off. Like how do you define agentic marketing? Like what does that term mean to you and the team over at Glean?
Katy Abbaszadeh – Glean
Yeah, I think we're all at different stages in our AI and agentic roadmap, so to speak. So I'm so glad you have this forum for folks to share where they're at and what they're doing with it. I think for us, the future vision and where we're starting to get a foothold is that agentic marketing really means that we're orchestrating autonomous and semi-autonomous agents to do work for us. So it's about grounding the agents in your own data, but then making sure it's measured on things like revenue impact. We as marketers really care about the outcomes, not just the inputs. So making sure that the agents that you have working on your behalf have an outcome that they're working towards. So I think about it as not just implementing a set of new age tools, but really a new operating model for how we plan work, execute work, staff work.
Sarah McConnell – Qualified
Yeah, that's such a good answer. As I think about my next question is like, how is it different than traditional marketing? And at the, guess the base of it, seems pretty obvious, like obviously talking to AI and agents and that's very different. But do you feel like agentic marketing is vastly different than traditional marketing, or is it at its base the same thing we've always been doing just with like a different twist to it? But how do you think about those two things differently?
Katy Abbaszadeh – Glean
It's a good question. I think more recently I've come to believe that it is a bigger shift from traditional. When it was first coming out, I was like, that's cute that AI can do that. But now I'm more on the lines of thinking, this is about moving from in traditional marketing where we were scaling humans with tools. Now we're taking agents that can scale outcomes with systems. So thinking about moving from activities to that outcome base. You'll hear me talk about that a lot. It's focusing on what the end result is. Channels to agentic chains, different agents that can do different things, and fully autonomously is the ultimate vision. I'll admit we're not quite there yet, but there's a lot of promise there. And then from individual ops, QAing on a human level, to governance, systems of governance that can kind of be the guardrails to what your agents are doing.
Sarah McConnell – Qualified
Yeah, and I really like that answer because I feel like in 2023, when it was first like generative AI, that was like, oh, this is cute. Like it helps me. It's a little, we used to call the co-pilot. So you're riding shotgun and helping me make my job a little bit easier, but it just didn't feel like this massive shift. It was just like a slight change. But then now here in 2025, I agree. I'm like, oh no, this is vastly different. And we are being asked as marketers to have a completely different mindset and framework because we now do have these autonomous agents that we need to orchestrate and we need to have them do full roles that we are holding them accountable to an output, and that does feel vastly different than what we used to do. So I think your answer is perfect in that sense.
Katy Abbaszadeh – Glean
Yeah, 100 percent. I would say even—not to cut you off—but as we're heading into next year planning, you know, I'm starting to ask my managers, when you're thinking about headcount for next year, how does AI play into that? Not to the full extent of like we're freezing hiring until you prove that AI can't do this, but really thinking about AI as part of your team. And I haven't created a prototype prototype yet, but, you know, what does the new org chart look like, and how do you highlight the agents that are doing work?
Sarah McConnell – Qualified
That's a good segue to my next question, which I'm so curious. Making that leap from experimentation with AI to truly leading with it, I think in the way you just spoke about it, you are truly thinking about this as, 2026 planning, how is AI scaling your team? How are you thinking about bringing them on to help already existing org structures? So that is a big jump. So what did it take over at Glean for you to feel like you're making that move from the like, this is cute, this co-pilot is helping, to like, okay, no. This is a part of our 2026 planning. We need to lead with agentic AI and agents.
Katy Abbaszadeh – Glean
Yeah, it's a good question. And frankly, I don't think it's going to just be one single leap. I think it's actually baby steps to get to that ultimate leap. Any new skill, there's a learning curve. We're all adopting this new thing in real time. It's changing, feels like daily, weekly, new stuff popping on the scene. So I definitely think we have to give ourselves grace. And I try to think about leading with vulnerability. Not everything we test is going to come out the way we want it to. Not everything is going to be a smashing success. So how do we think about, as leaders, rolling up our sleeves, getting in the mix, being on the vendor calls, you know, workflow write-outs of jobs to be done, where can AI fit in? I think that's a really important piece of it. And I meant to introduce Glean lightly at the start, just because I think it helps to have some context of what Glean is. We're a work AI platform. So I have the benefit of working for an organization that is trying to transform work with AI across the board. So we're a horizontal platform that connects all of your apps and data at work in a safe way. So I have a bit of a cheat code in that my product that I'm marketing actually helps me be an AI-first marketer. But back to your question about what does that leap—I think the practical answer I'll give is I really recommend starting with one high-leverage workflow that you can really easily map out from what you're doing today and then kind of build in what are those things that can be replaced by AI. So an example is an ABM activation. Obviously, we hear a lot about content generation for AI being a major use case, personalization. So—
Sarah McConnell – Qualified
You're good.
Katy Abbaszadeh – Glean
—what is one ABM activation you're already doing that you can now kind of shift and think about it in an AI framework.
Sarah McConnell – Qualified
Yeah, I think that you talked about Glean being an AI-first company and giving you sort of that AI-first mindset, and obviously working at Qualified, I feel the same way. We have an AI-native product. We have our AI SDR, Piper, which we'll talk about, but it is such a unique position to be in because it does give you sort of the permission to really lead with that AI-first mentality, which I don't think is necessarily the case at every single organization. So I think to your point of starting there and it's starting with that experimentation, but picking like one thing, but being very lucky to work for organizations that do—
Katy Abbaszadeh – Glean
See you on the next video.
Sarah McConnell – Qualified
I'm assuming like your leadership team is asking you, like, where is AI showing up in your day-to-day because you're an AI product? So you're saying help to that accountability—
Katy Abbaszadeh – Glean
Yep. Yeah, absolutely. It takes forcing yourself to think about it every day. And I think at first, like we talked about it, it's so cute. It's fun. It's an experiment, right? But now, still maintaining time and space for the team to think about it, get creative with it, test and try and fail. I think the failures are really important in these baby steps, ultimate leap. So constantly trying to keep it top of mind.
Sarah McConnell – Qualified
Absolutely. Now, I'm curious, you said you oversee DemandGen, you oversee events and ops. Where is agentic AI showing up the most visibly on your team, whether that be from a DemandGen perspective or across your largest, like the marketing org as a whole? I think listeners are always so curious of, okay, agentic AI, agentic marketing sounds really cool, but I don't know where to put this into play. So where is it showing up for you on your Glean team most prominently?
Katy Abbaszadeh – Glean
Yeah, I think generally it's in those outcomes we talked about earlier, what we're kind of referring to as revenue moments. So you'll hear it come up when we get to the Piper part of the discussion, but that inbound speed to lead piece is really important to us. Outbound sequencing, some of those ABM moments where we can help the human SDRs that we do have be more personalized and proactive. Content generation and personalization we talked about already. Pre-meeting research for our reps that we as marketers can kind of curate for them. And then that, back to the inbound speed to lead, that 24/7 coverage that we just don't have to worry somebody's not going to get that custom touch that we want them to have. Obviously, productivity is a huge win across the board. So just planning time for campaign output.
Sarah McConnell – Qualified
Them.
Katy Abbaszadeh – Glean
You know, we've seen some of our planning cycles go from something like 15 days to a day because we can feed an agent the ideas, brainstorm with it, have a plan right then and there. And then more specifically within my team, I would say our digital marketing team is really leaning in heavily. So from a content generation perspective, we've got our SEO strategy really humming with AI workflows. Obviously, Piper is a big piece of that, our AI SDR on the website. And then even getting into some vibe coding on new website modules. So if you check out our homepage, you can see some that were vibe coded and then kind of built into the glean.com homepage. So some fun stuff there. And then the ABM stuff is a big play as well.
Sarah McConnell – Qualified
That's amazing. It's so cool to hear how many different aspects of your organization are adopting this, because I feel like getting it into one part of the organization always feels like that first step, and that can be not easy, but is easier to do, but getting it to be sort of an org-wide mindset, having been on the forefront of that, is much, much harder. And then my last question of this section of our conversation today, I'm very curious, do you have a hot take about AI and marketing that most people would not agree with you on?
Katy Abbaszadeh – Glean
Oof, hot takes. Love a hot take. I have to go back to the cutesy theme of how it burst onto the scene. I think right now most teams are overestimating the novelty and underestimating the operations portion of it, the backend portion of it. Agents are ready, and I'm not saying that Glean has it down perfectly just yet. They're still learning. But agents are ready, and it's really about—
Sarah McConnell – Qualified
Hahaha!
Sarah McConnell – Qualified
Them.
Katy Abbaszadeh – Glean
—org design, underlying data governance, how you're measuring things that become the bottlenecks to success. So I think the winners of this wave, so to speak, will really treat this as a go-to-market transformation and not just spot tools that you're rolling out here and there, but really rethinking how you plan, staff, and actually execute the work.
Sarah McConnell – Qualified
That's such a great answer because I do think it is a hot take because I do think we're hearing more in the market right now of people being like, well, maybe it's not as impactful as we thought it was going to be. And I worry that it's going to make marketers kind of rest on their laurels a little bit and be like, well, you know, if other people aren't finding success with it or I'm saying, you know, that's not impacting our bottom line, I don't want to invest the time and the effort to get up to speed with agentic AI and how I'm going to bring it onto my team. And I worry that with those two shifting narratives of like, no, it is really important versus like, OK, maybe there's not as much impact of it is that too many marketers will wait and say like, it isn't going to be that impactful. And then when it is, we'll feel left behind. So I love that your take is like, no, it's here. It's ready. It's kind of on us to build that infrastructure into our team. And I do, I agree with you. I think the teams that are embracing it early and are building in that infrastructure are the ones that are going to win and be way ahead. All of we're like, no, this is here to stay. This is going to be a vital part of our organizations. Encouraging other marketers to stay up to date with it and stay up to speed I think is so, so important.
Katy Abbaszadeh – Glean
Yeah. And I just want to say I understand the feeling of maybe burnout or, I tried this thing that I thought was going to be so great and I hit a wall and why would I try it again? I think even just having the ideas for, you know, certain workflows or what will be life changing when AI meets you there, having that scaffolding ready to go when AI catches up, you're that much further ahead. So. Keep it up.
Sarah McConnell – Qualified
Totally. Okay, so shifting more gears into what you guys are doing over at Glean, you guys have hired Piper, the AI SDR from Qualified, which is amazing. We're so happy to have you guys as customers. You've renamed her Taylor. What was it about having an AI SDR agent that made you and the team lean in? Like, what was it about that particular use case that the Glean team was like, okay, this is something I think we can start with or we can start utilizing within our organization?
Katy Abbaszadeh – Glean
Yeah, I think for us, as I mentioned at the top, all roads really lead back to PipeGen. It's the North Star metric for my entire team. Even Marketing Ops is constantly refreshing the dial on the Salesforce dashboard, even though they're not the ones with the exact ROI target. But the digital team in particular, who owns the website, really saw that pipeline was being left on the table because we couldn't rely on—
Sarah McConnell – Qualified
Mm-hmm.
Sarah McConnell – Qualified
What the—
Katy Abbaszadeh – Glean
—couldn't ask our human SDRs to be online 24/7. And so that team worked really closely with our Qualified CSM, Sean—shoutout—to put together a really compelling case to hire Piper. She is fully available on the website. We don't actually ask any of our human SDRs to be on that channel any longer. But I will say we don't plan on fully replacing our human SDRs. We really believe that there's an important place for them as well. But it was about pipeline. We didn't want to leave any dollar left behind.
Sarah McConnell – Qualified
I totally agree with you. And I feel like we went through the same sort of transformation here. I obviously get to be like on the forefront of our products and be our guinea pig, which is a blessing and a curse. We're the same way, and like asking our reps to be online all the time. Obviously one, can't—like they have to sleep, they have to eat. They cannot be online all the time. And even when they were, I just was just talking to someone else about tire deflection and like these, or sorry, tire kickers and like kind of deflection of people on the website is other, I'm like, am I really utilizing my SDR teams the right way when I'm seeing these conversations they're having with these people that aren't really that interested in like, yes, we can deflect them through the product. But I'm like, what if I—if I just have an AI agent do all of this, they can go focus on the things. And this is such a played out term, but like they can go focus on the things I want them to focus on and a lot more high-value things. Like cold outbounding is hard. It's a very hard thing to do. It's hard to do really well. We do have an outbound AI SDR that we use as well. But when it comes to like named accounts and really personalized experiences, I still very much believe in a human touch for that. So I'm like, if I can free up their time to do the good research and actually write really compelling messaging, is their time better spent doing that? So yeah, I agree with you on that. It's not about replacing them, but like, can we find areas for them that are gonna be more impactful for organization and for their career trajectory? Because responding to like mundane—
Katy Abbaszadeh – Glean
Thank you.
Sarah McConnell – Qualified
—answers on our website all day long, it's not really a great learning experience for them necessarily.
Katy Abbaszadeh – Glean
100%. And we still arm the human SDRs with their own AI for that personalization content generation. But I think another played-out term, human in the loop, is still really critical to success across the board.
Sarah McConnell – Qualified
Yeah.
Sarah McConnell – Qualified
And now beyond pipe generation, was there any unexpected things that bringing on an AI SDR agent helped unlock for your team? Is there anything you noticed that went beyond your expectations of pipeline generation and following up with leads?
Katy Abbaszadeh – Glean
Yeah, well, you actually just touched on it. And perhaps it shouldn't have been unexpected. But I do think the biggest eye-opener for me personally was the DQ rate. Piper does such a good job of weeding out prospects who just aren't a good fit, maybe yet or never will be. And so that saving for the human SDRs’ time was huge. So I think that was an eye-opener. And then just other metrics that we've been keeping a very close eye on with related to Piper is that time to first touch, you know, compared to the control of when the human SDRs were on when they could be, DQ rate, as I mentioned, and then meeting acceptance and show rate. Like, is there a difference between what Piper's able to book versus our SDR team? And then incremental meetings pipeline obviously is that outcome North Star that we're going for, and we've actually seen an uptick in that, I think due to the speed and efficiency, and she's always there.
Sarah McConnell – Qualified
That's incredible. You kind of already answered this, but I want to dig in a little bit deeper. As a marketing leader that's obviously taking AI very seriously within your organization, if you were talking to a peer and they're like, I just don't know where to start. I feel very overwhelmed. I think we've all felt that way at some point. What advice would you give to them on where to start? What is the best starting point to get over that hump of like, okay, I've run into too many walls with AI. I don't feel like this is worth my time. What advice would you give them?
Katy Abbaszadeh – Glean
This is a tough one because I think every organization is a little bit different, and depending on what your ultimate metric is, you want to make sure that your stake in the ground with AI has an impact on that. So I guess I would say, sort of back to the high-leverage use case, pick something that is a money-in, money-out workflow that you know like the back of your hand in the traditional way of marketing, and really try to map that out as specific jobs to be done and then rework that as an AI flow and do a comparison. Run a six-week A-B test and see what you get at the end. I think being clear about what you are measuring in that experiment is important, but then kind of let it do its thing and—what happens. I think that's the best way to really dip your toes in, and maybe in some cases traditional wins out because your underlying infrastructure isn't quite ready yet or, you know, I would say even if AI doesn't win the first battle, don't give up because there are always things that you can continue to refine and test.
Sarah McConnell – Qualified
And you said it earlier, I think it's okay to fail. I think that is something as marketers—don't, maybe it's just the circles that I run in—but marketers tend to be pretty type A for the most part. And I think failing is not something we love to do. I think it's a demand gen thing. I will say we tend to be cut from the same cloth. But failing is hard. Like I think if, as a demand gen person or someone that owns a pipeline number, you tend to be a little competitive. You've got big numbers that you're always looking to hit. You've got that drive in you, and failing, even if it is something that's new and different—
Katy Abbaszadeh – Glean
Guilty.
Katy Abbaszadeh – Glean
Yeah.
Sarah McConnell – Qualified
—can feel very frustrating. So I like that you said that early about be okay failing, but if you set the outcomes that you're looking for, it makes it a lot easier to fail fast and say like, okay, this isn't working, let's pivot, let's move on versus just spinning your wheels on something for a long time that ends up not playing out and/or giving up too fast. Like it might be something that you want to stick to. So I think your advice is great and also like fail quick and move on.
Katy Abbaszadeh – Glean
Yeah, I was just at an AI for Women meetup in Minneapolis last week, moderated a panel there. And I think that was my number one takeaway. It was really, we have to be able to embrace failure right now. And this wasn't marketing specific. It was kind of across the board—sales enablement, IT folks. But it's so new and changing so fast that if you expect it to be perfect and intuitive and something you can just roll out and walk away from—
Sarah McConnell – Qualified
Yeah.
Katy Abbaszadeh – Glean
—then you're definitely setting yourself up for failure. But if you take the bite-size failures along the way, like you said, fail fast and learn from it, then you're setting yourselves up—and your team—for success.
Sarah McConnell – Qualified
Absolutely. Now, as we think about AI in general, it's obviously made a massive impact on us as marketers and within our organizations. But I think it's also making a pretty big impact on the expectations of buyers. So do you have any advice for us as marketers? Like, how do you think AI is changing buyer expectations? And are we underestimating anything that we need to be cognizant of that AI is really changing for our buyers that we need to keep up with?
Katy Abbaszadeh – Glean
I'm gonna steal an answer from a CEO of a vendor I chatted with a few months ago, but it has just stuck with me. Eventually, you can imagine a world where we're marketing not to humans anymore, but to agents who bring back answers from queries on ChatGBT. Those agents are gonna go out to the websites, figure out how to compare different products and—
Sarah McConnell – Qualified
Mmm.
Katy Abbaszadeh – Glean
—serve up an answer where buyers expect a one-minute internet immediate answer—what's the best choice? And so I think we as marketers really need to be aware that some of our key metrics are going to shift. Should we care as much about website visitors in the future? Because it might just be an agent pinging your website or having it cached. So I don't know. It's a bit of a scary frontier because we're—
Sarah McConnell – Qualified
Yeah.
Katy Abbaszadeh – Glean
—we love our websites, don't we? But I do think we have to think about how our metrics shift, who we're marketing to shifts. So a lot to think about on that front.
Sarah McConnell – Qualified
I just had someone on the podcast last week and they kind of shared the same sentiment. And I think that's been like prolific around the marketing—just any peers I've talked to is like, yeah, website traffic is down. Like I think LLMs and the way that people are engaging with search is just changing so drastically, so quickly. But the continued sentiment has been the traffic that is coming from those, the AI sources, tends to be a much higher quality because they've already furthered themselves so much farther down the buying cycle, because of the way they're engaging with them. And at some point it's gonna be agents coming in, querying your website, and they're gonna come back with an answer as the buying cycles go faster. There are much higher conversion rates of those visitors coming from those sources converting into opportunities, and then once they're opportunities, they're closing quicker. As marketers, it will be so—it'll be a new challenge to convince our executive teams that like, yeah, it's okay that website traffic is dropping because the outputs and the metrics that we care about are still—as long as you're maintaining those—it's fine. But it is very scary to suddenly see this dramatic drop in traffic and sources. That's something I think we've all been struggling with just based on the conversations that I've been having.
Katy Abbaszadeh – Glean
Yeah, totally agree with that. And while the traffic from the LLMs is still small compared to everything else, to your point, the quality and the conversion and the speed is so much better. I think we just have to continue focusing on outcomes and that ultimate outcome that we're looking for versus the input metrics that we've typically kind of hung our hats on. So I think it makes it that much more important to drive that alignment with sales. What is the definition of pipeline? At what point, at what stage in the cycle are we measuring it? What coverage do we need? And making sure that we as marketers are speaking that same language so that we're not over here saying, the traffic, but the MQL.
Sarah McConnell – Qualified
Hahaha. Now, on that same note, like we just talked about, obviously search and web traffic has been a huge change, but is there another big—like, what would you say is gonna be the single biggest change to B2B marketing in the next 12 months, or is it going to be search in the way that people are interacting with our website?
Katy Abbaszadeh – Glean
Yeah, predicting the future. Yes, here we go. You know, I think this is less in my wheelhouse from a demand perspective, but as a buyer myself, I'm feeling like this shift is something we can't ignore. Pricing structure and pricing expectations. People do not want to get locked into—
Sarah McConnell – Qualified
I know, I'm like, hey, can you get out your crystal ball and just like tell me what the future holds, please?
Sarah McConnell – Qualified
Mm-hmm.
Sarah McConnell – Qualified
Yeah.
Katy Abbaszadeh – Glean
—annual contracts because things are changing so fast and you want to be able to test. So I think it's maybe a little bit easier to price that way in a consumer market, but on the B2B side, we have to be thinking about how we shift to that expectation so that people can be confident, you know, testing and using a product without getting locked in for one- to two-year contracts.
Sarah McConnell – Qualified
Yeah, I think that's a really good one. And earlier we've kind of talked—I think we've touched a lot throughout this conversation on use cases for AI agents. Like, I think you gave some really great examples that you're seeing within your own marketing organization. But one of the things I'm so curious about is what do you hope to see more of soon? Like, are there any areas within your marketing team that you're like, man, I really wish there was some sort of agentic solution that could help me scale this that you're just not seeing as much of in the market right now?
Katy Abbaszadeh – Glean
Yeah, this one's maybe less marketing focused, but I really am liking the agents that are more focused on coaching. So, executive coaching—I'm, you know, moving up in my career. How can I make sure I'm speaking to the right audience and the right tone and saying the right things? For our sales team, first pitch coaching. So I think that will probably become more formalized. It's a really interesting use case. One that's not necessarily a use case, but I do think we need to get there as an industry is open metrics standards, so that we can actually compare agent performance apples to apples. Obviously, each company is going to have a tech stack, different way of building agents. But if there's a way that we can—
Sarah McConnell – Qualified
Hmm.
Katy Abbaszadeh – Glean
—as marketers, always asking, OK, what's the industry average for CAC and web visits and all of that? We need a similar way to benchmark how we're doing in AI, and I don't think we're quite there yet.
Sarah McConnell – Qualified
That's probably my favorite answer I've heard today. And as you said, industry benchmarks, I like had a shiver run through me because I think in demand gen, the number of times we're like, well, what's the industry after—I'm like, I don't, I don't know. Like it's so different across organizations. Like in every company I've worked at, it's been so different. I tried to crowdsource recently, like as website traffic has been dropping, I'm like, where are people seeing—like what's your percentage breakdowns of web traffic? And it was—when I say all over the board is an understatement. So being able to have more standardized averages, especially as we think about AI, like, I like jump for joy. I was like, yes, can someone please make this if you're listening, because we just don't have that insight. And I feel like we're always flying a little bit blind.
Katy Abbaszadeh – Glean
Yeah, well hopefully the LLMs can already help because they've got access to all of the information. So you can get more specific on what do we even mean by industry when we say industry average. So I think that's promising, but we gotta do some work to make sure it's there for everyone.
Sarah McConnell – Qualified
Totally. Okay, my last question before we get into my favorite section, which is the lightning round. I'm curious if there is something that you think most B2B orgs will still get wrong about AI a year from now. So before I ask you—you know, get out your crystal ball, pull it back out—what do you think we're getting wrong now that you don't think is gonna get better? Or that, I guess I want to phrase that is, what do you think we should be focusing on because you're worried it's a problem or that we're still getting wrong?
Katy Abbaszadeh – Glean
Yeah, I'll do a callback to one that I kind of mentioned earlier, but I think it's about focusing too much or celebrating too much the demoable moments, the like little delightful assistance that like, hey, look how cute this is, and ignoring what that boring backbone takes to make it work more in a complex manner. So the data governance—at that breakfast I was at last week, there was a ton of talk about changing the terminology of data governance to AI governance. And how do we—you know, another phrase we like to use, garbage in, garbage out—how do we make sure that the AI is set up for success? And building that foundation is hard. And there's a lot of question on who owns building that foundation. You know, when you're in marketing and—
Sarah McConnell – Qualified
Mm-hmm.
Katy Abbaszadeh – Glean
—you're partnering with rev ops or IT, you know, who really owns that and how do you make sure that's set up appropriately? I think that'll be the tough thing for people to get right. So I'm interested to see how we solve that problem collectively.
Sarah McConnell – Qualified
Yeah, that is a great answer. Okay, I love to wrap up these episodes with our lightning round. So it's quick questions with quick answers. So to get started, other than ChatGPT, what was one of the first AI tools that you experimented with?
Katy Abbaszadeh – Glean
This is going to sound really lame. I promise I'm not going to sell Glean. But it was Glean because I was just coming back from maternity leave right after ChatGBT exploded onto the scene. And our engineering team had just done a hackathon. And three weeks after the hackathon, we launched Glean Chat. And so I really played with ChatGBT through Glean before I even played directly with ChatGBT.
Sarah McConnell – Qualified
That is such a—what a wild thing to come back to maternity leave from is like, I get back from maternity leave in August and I was like, what is vibe coding? Like, can someone—that's been four months and I feel like suddenly it was a completely different world. And that was the least dirty day. I feel like you came back to like a whole—literally a whole new world. Okay, what do you think is the most overrated buzzword in martech right now?
Katy Abbaszadeh – Glean
Right? Exact same experience. Yes, it was insane. I'm gonna go with personalization at scale. And I think this one is one that we like to use. It sounds great. We all talk about ABM. But without that governance piece and a measurement on outcome, it's just louder spray and pray. So.
Sarah McConnell – Qualified
It's so true. Okay, who is a marketer that you think is ahead of the curve on AI that listeners should go follow on LinkedIn right now?
Katy Abbaszadeh – Glean
OK, I will give you a name, but I first want to start with a shoutout to another podcast that I recently found. So on top of this podcast, which obviously you're already following because you're listening to this, Humans of MarTech has been a really great one that I've recently discovered. So shoutout to them, curating a ton of really interesting content from marketing leaders. And then, for an individual person, I have to give a shoutout to our head of digital marketing, Daniel Henderson. He has been partnering closely with Kevin Indig, who is somewhat of an AISCO influencer out there. And he's been involved in a lot of our AISCO projects. He's got the scrappy, rapid testing approach that has really helped us kind of frame our thinking and where we want to place bets. So Kevin Indig is a good one to check out.
Sarah McConnell – Qualified
Amazing. Okay, last question. If you could automate one part of your life with AI that's outside of work, what would it be?
Katy Abbaszadeh – Glean
I would absolutely love an AI stylist or personal shopper. And I'm sure there's a version of this that probably already exists, or I could go train my own custom GPT, but who has the time for that?
Sarah McConnell – Qualified
Absolutely. So I agree with that. Okay, Katy, I appreciate you joining us so much. Thank you for joining The Agentic Marketer podcast. It was great having you on and having you your insights. So thank you so much.
Katy Abbaszadeh – Glean
Thank you, Sarah.
Discover how we can help you convert more prospects into pipeline–right from your website.
.svg)
