Customer Zero: The Story Behind Grammarly's Boldest AI Experiment

Episode
5
Jul 2, 2025
Listen on Spotify buttonListen on Apple Podcasts buttonWatch & Listen on YouTube button
Show Notes

In this episode of AI for GTM, Emily Sandison, Head of Demand Generation at Grammarly shares how a bold campaign put their AI capabilities on display, and made transparency the headline.

00:00 Introduction and Guest Introduction
00:42 Grammarly's Recent Changes and AI Integration
01:09 AI Tools and Best Practices in Marketing
03:31 Signal Overload and AI Solutions
07:24 AI in Content Creation and Marketing Strategy
14:24 Customer Zero Campaign Insights
19:39 Future of AI in Marketing and Hiring
23:32 Exciting AI Developments and Final Thoughts

Transcript

[00:00:00] Emir Atli: Hello everyone. Welcome back to AI for go-to-market podcast. This is episode five. I'm with Emily Sandison from Grammarly. Emily, thank you so much for making this. 

[00:00:07] Emily Sandison: It's good to be here. 

[00:00:09] Emir Atli: Thank you. Do you wanna do a quick intro? 

[00:00:10] Emily Sandison: Yeah, I'd love to. Yeah. So Emily Sandison, I'm the head of Demand Gen for Grammarly and, have been there for about a year leading our enterprise team and a group of demand gen marketers.

[00:00:23] Emily Sandison: email marketers, and some content marketers. So it's a great little ride. 

[00:00:28] Emir Atli: Awesome. How big is a team? 

[00:00:29] Emily Sandison: Across our entire marketing function. We're well over 50. Specifically kind of within demand gen, we're a team of about, eight to 10 mighty marketers. 

[00:00:41] Emir Atli: Nice. Awesome. Well, Grammarly just had massive news. Yeah. Congrats on that. Thank you. I saw you acquired coda a couple months ago, right? 

[00:00:50] Emily Sandison: Yeah. Yeah. It's been a huge time of, of change and transition. It's really exciting. Um, you know, an exciting time in the, in the space in general. 

[00:00:57] Emir Atli: Yeah. Awesome. Well, I know there are [00:01:00] lots of changes in Grammarly as a business, but also there are lots of changes in marketing and go-to- market, especially around AI.

[00:01:05] Emir Atli: I would love to just. Dive into it. And I would love to know how you're using AI day to day. How are you thinking about as a demand gen leader? 

[00:01:12] Emily Sandison: Yeah, I mean, it's essential, right? Everyone's saying that. Um, I think we have done a good job as an organization at this stage of maturity of AI, of giving people the ability and the flexibility to use the tools that are actually working for them.

[00:01:26] Emily Sandison: Um, so we have a wide variety of tools that folks are using and I think. It's incredibly important to help folks find creativity with it, find what's working, what's not. Um, so we've been doing that a lot, sharing a lot of best practices that we're finding, sharing where it completely falls flat on its face.

[00:01:44] Emily Sandison: Um, and I think it's a really interesting time of like, probably way too many AI tools than any of us could even use. And, you know, that will lead to consolidation at some point. Mm-hmm. But at this point, I think the more, the merrier a little bit as we all get comfortable with it. Um, we actually [00:02:00] launched a, a little quiz, a fun quiz, um, on AI maturity this week.

[00:02:03] Emily Sandison: And, uh. So you can take the quiz and you get a little animal icon at the end. So it's been fun in Slack this week as everyone's kind of sharing where they are on the journey. Mm-hmm. Um, and that's, that's really humanized it, I think. Um, and opened people up to sharing where they are. And so, you know, as a team leader kind of saying, alright, here's what the team looks like.

[00:02:20] Emily Sandison: Um, where can we go from here? 

[00:02:22] Emir Atli: Is it like an internal test or is that external? No, 

[00:02:24] Emily Sandison: it's live. Um, yeah, so on our website, um, I think it's been shared on LinkedIn quite a bit, um, and. Get anywhere from a, you know, an ostrich with their head in the ground to a, an octopus, I think is, is the most developed. Nice.

[00:02:37] Emily Sandison: So, um, it's super fun. 

[00:02:38] Emir Atli: Yeah. Octopus is a good animal for maturity. Yeah. Yeah. Um, in terms of like specific use cases or during the day, how, how do you want your team to use AI or what are some use cases that you find valuable with AI? 

[00:02:52] Emily Sandison: Yeah. So I think for marketers in particular, you know, there's been like the natural fit of like, [00:03:00] hey, you know, AI can write all of your copy.

[00:03:01] Emily Sandison: You don't ever have to do anything again on the content side and. Like, that's just not true. Um, and so, mm-hmm. I think I've actually found a lot more use for it internally. Um, so rather than kind of copy that sees the light of the day or messaging that sees the light of the day, I'm using it a lot for internal, um, communications.

[00:03:21] Emily Sandison: Obviously, you know, I. Grammarly is side by side on that one. Mm-hmm. Um, but I'm also using it to summarize a lot of information and data, which is speeding a lot of things up. So, you know, in particular, um, we as an organization have kind of moved from zero to 110 in terms of signals in the past six months.

[00:03:42] Emily Sandison: Um, so we turned on some ABM platforms, um, and we actually also built our own internal, um, AI model that looks at product signals, um, and how we kind of bring all of that together for the sales team and go to market teams at Grammarly and, um. Therefore we have just like, kind of dumped a [00:04:00] bunch of signals on people and I think there can be a lot of overwhelm.

[00:04:03] Emily Sandison: Um, I find it fascinating. Mm-hmm. But it's a lot to wade through and so, um, the team has been taking a lot of those signals and actually slowing down to speed up a little bit. Um, and so looking to find what are maybe some unique trends within specifically us, um, for success for closed won, and how can we actually replicate that, um, and kind of reduce some of the noise there to actually pick like three to four things that actually matter, um, and run that as a sales play and for the sales team.

[00:04:34] Emily Sandison: Um, because I think, you know, without taking the time to do that, you end up going to market off of signals like, hey, Grammarly just raised a bunch of money. Um, I'm, you know, I'm getting emails left right and get center from that. And it's like, I think everybody's doing that. So I actually listened to Tito's episode recently and I thought that was really interesting to say like, you actually have to be incredibly unique, and so I think on the marketing [00:05:00] side of things, um. Bringing clarity there and bringing it down to a number of signals that are dependable and that are like three to five, so that a human can actually grapple with what's going on versus looking at an account and being like, I mean, there's like 17,000 lines of activity here.

[00:05:18] Emily Sandison: How do I make sense of any of it? Um, so I've been doing a lot with that and just summarizing and synthesizing that not only to kind of create a little bit more of a playbook there, um, but also to celebrate wins too, where it's like in the past, to celebrate something that came from marketing, um, and, and get that whole path all the way to conversion.

[00:05:37] Emily Sandison: Um. You know, it took a lot of digging. And so to be able to kind of do that succinctly, um, and read that out to the team has been really great. 

[00:05:45] Emir Atli: Yeah. Um, so you are at, like, you're tracking hundreds of signals right now, and you wanna get to a point where you're, I don't know, three or four signals that make sense?

[00:05:52] Emir Atli: Is that, is that correct or are you at the point where you only do three or four signals that make the most sense for Grammarly. 

[00:05:58] Emily Sandison: Yeah, we are at [00:06:00] hundreds at this point. Um, and so I think getting down to those three to five, working through those a little bit, they're gonna change over time. Yeah, they should.

[00:06:07] Emily Sandison: Um, and then expanding that back out as the team's kind of ready and up to speed there. Um, because I do think the number just leads to decision paralysis. 

[00:06:17] Emir Atli: Mm-hmm. How do you deal with that? Like signal overload, signal paralysis, I'm sure you are going through that as well with the team. Are there any best practices that you found, um, to be useful?

[00:06:28] Emily Sandison: Yeah. Um, I mean, honestly, this is where like AI has been really, really helpful is just helping bucket them. Mm-hmm. So it's like you're not looking line by line of activity. You're getting really good buckets of like. You know, there's a bunch of visitors and people spending time here on your website. Um, and I know Hockey Stack can visualize all this as well and so, um, just I think those buckets are really helpful. Um, knowing some of the down funnel signals that you can get from like third party sites, um, you know, it's like, do those matter for [00:07:00] us in particular? And is that a real sign of, um, you know, timeliness to reach out? Um. And then we have product signals all day long as well.

[00:07:09] Emily Sandison: But um, across. You know, folks actually using the product, and so figuring out how we bucket those sorts of things as well, and kind of combining them together clearly to say, okay, we have A plus B plus C, you know, now is the time to go reach out to this account. 

[00:07:24] Emir Atli: Yeah, I mean, Grammarly is a massive PLG business.

[00:07:26] Emir Atli: It's probably exponentially harder than like a pure sales lead motion to bring all of those signals and action them. 

[00:07:32] Emily Sandison: Yeah. Yeah. And we've, you know, kind of had to crack that nut a little bit by having some external tools on the ABM side and then building our own internal tool internally. There was really no other way to, to go about that.

[00:07:44] Emily Sandison: And so it's been really cool as that's been just an interesting motion to help the sales team understand what's happening across enterprise free, et cetera, like every single user type and, and what might be important there. 

[00:07:57] Emir Atli: Yeah, I mean Grammarly is probably in a ton [00:08:00] of companies' target account list, top 50, top hundred accounts.

[00:08:03] Emir Atli: And you mentioned that you are building stuff internally as well with AI making things really easy to build internally. Do you see a shift in like your budgets in terms of tech versus like building stuff internally for Grammarly? 

[00:08:16] Emily Sandison: I think so, and I think, you know, we've taken the approach similar to the approach of tools for the team and just kind of having a number and seeing what, you know, what resonates, what gets used? Um, we've taken that approach right now with software and so there are definitely areas in which our A BM tools and our internal tools overlap a little bit, and definitely areas where we need both, right? And so letting them both run for a period of time, um, before kind of making that cut or, or finding what we can learn from one, um, and consolidating so.

[00:08:48] Emily Sandison: Same with AI. I think on all of the signals too, like we're gonna get to a place where you, you know, feel like you have a really surefooted set there. 

[00:08:56] Emir Atli: Do you feel like you'll buy less tools in the next few years? [00:09:00] 

[00:09:01] Emily Sandison: I think so. I don't know. It's so hard to say, but I, I mean, I do think so, and I think right now there's just so many things for everything.

[00:09:09] Emily Sandison: And on the AI side, right? Like. The main tools will all of a sudden have their own parts and pieces, and I think that will kind of reduce some of the noise. 

[00:09:17] Emir Atli: Mm-hmm. And you said something very interesting, um, you said basically you don't want your team or, or yourself to create content with AI. Um, is it like not creating the entire, like content or is it just not using AI completely in the content or, 

[00:09:34] Emily Sandison: Yeah, I think it's more so, not copy and pasting. Um, in terms of, I think we'll get to a place right as AI's learning context and learning, um, personality a little bit. Like, I think we'll actually get to the place where you could kind of say, alright, press go and post to LinkedIn. Let's just say, um, I don't think we're quite there yet.

[00:09:56] Emily Sandison: Mm-hmm. Or maybe that's just my hesitation as a control [00:10:00] freak to be like, I'm gonna read over this and make sure that this sounds like me. Um. I think we're getting closer. It was actually really cool. We, um, a recent campaign we launched specifically for the marketing persona. We call it Customer Zero. Um, it's actually us as marketers at Grammarly.

[00:10:18] Emily Sandison: Writing a marketing campaign for Grammarly. And so we were really forward about it, of like, hey, Grammarly helped us write this landing page and all of the emails and all of the ads that we're pushing out as a part of this campaign, both from ideation all the way to Final Touch. And um, part of that, kind of the strategy behind that was all of us, um, sharing it on our own LinkedIn and our own social media.

[00:10:41] Emily Sandison: And it was really cool 'cause we were all using AI in the writing, but it did all sound like our unique tones and voices. And so I think, um, I mean there's, it's just leaps and bounds even from three to six months ago, um, in terms of the personalization. But I definitely don't think, you know, I would set up [00:11:00] a write these social posts and press send without kind of that human review layer right now. 

[00:11:05] Emir Atli: Mm-hmm. And how, I mean, marketing has always been like a lot of art, a lot of data as well. But I'd say it's a creative function, creative department. I'm not sure how many like B 2 B companies are actually created with their marketing, but marketing, by definition should be creative. If AI is gonna be able to do a lot of the things, and I mean, right now it's on its current form it's not creative, it's, it's basically pattern matching across a lot of different samples. How do you see the role of marketing changing over time? 

[00:11:33] Emily Sandison: Yeah, I mean, I think it's, it's gonna be an augmentation, right?

[00:11:38] Emily Sandison: And so there will be. Jobs to be done that AI can completely do. Um, that's why I think, like even right now there's just a lot of, some of the data processing and like internal communications that totally speed up mm-hmm my workflow and make me more efficient, I think as, as it kind of gains context and personality, um, [00:12:00] it will just make us be, I mean, ultimately faster. Right. And you're still gonna need that creativity and personality because. Uh, you know what? A colleague of mine who is maybe wittier than I prompts to AI is gonna get a different output than I am. And so, um, yeah, I was thinking about that actually earlier this morning. Just like, hmm, I think we're gonna get to a place where, you know, prompts even are as dry as they are today and there's some more personality integrated in them.

[00:12:28] Emily Sandison: And I think that will kind of set things apart where, um, that creativity will still shine. 

[00:12:33] Emir Atli: Mm-hmm. Do you see any, I don't know, demand, gen A BM, field marketing, any of those or more marketing departments getting more or less, um, attention or getting smaller or bigger? 

[00:12:46] Emily Sandison: That's a great question. Um, I think that, I think we're gonna become maybe a little more generalists. Mm-hmm. Um, in that, like some of the very specific tasks, AI can take some of that. [00:13:00] Um, and so we'll be able to oversee a little bit more, um, because some of those really, really specific things are taken off the plate. I do actually think it's really interesting right now.

[00:13:11] Emily Sandison: Um. Especially for events and field marketing, um, how important those are. I think, you know, people are craving that human connection. Yeah. Um, and so we are seeing a ton of success with our field events, um, and our field marketing program, um, you know, as folks are kind of really desiring those in-person connections and content.

[00:13:35] Emily Sandison: So I think as we continue to speed up in this one way, you know, we're still gonna find we're all humans and we wanna connect. And I think, um, it's almost like a course correction maybe from 2020 where all events stopped to actually events being mm-hmm. Being virtual. Super. Yeah. Super important. 

[00:13:50] Emir Atli: Yeah. Yeah.

[00:13:52] Emir Atli: Yeah. I see that. Um. We actually just had a Giants game last night, uh, where we host a lot of customers and partners. That was great. Yeah. [00:14:00] Everyone wants, I mean, I don't know if people wanna work in an office, but they definitely want to go to events in person versus like looking at a Zoom screen for webinars or anything like that.

[00:14:10] Emily Sandison: Yeah. And I imagine people were excited to talk with their peers, right? Mm-hmm. And like get together in a forum and actually hear what's happening in the trenches and, um. Yeah, so I I love that. I'm glad you're having success there too. Yeah. 

[00:14:24] Emir Atli: Do you wanna talk a little bit more about the Customer Zero campaign and maybe how did you plan it?

[00:14:29] Emir Atli: What did you learn? What were the metrics or results that you wanna share? 

[00:14:33] Emily Sandison: Yeah, so we started thinking about this, um, I wanna say in like December. It was like, I, I think it's the time. Like I think we're ready both as a team and, um, you know, as a technology that we could actually do a whole campaign in Grammarly.

[00:14:51] Emily Sandison: Um, and we've never been that forward of like, Hey, we've actually used Grammarly in all of this. Um, and so, um, we really put it together [00:15:00] and, you know, from a demand gen and like a revenue marketing perspective, I also was like. This is a bit of a, a risk, it's more of an awareness play. Yeah. Um, you know, it's not a typical, uh, webinar or gated asset in any sort of way.

[00:15:13] Emily Sandison: And, um, to push the boundaries on that I think is really important. So, um, I. Yeah, so we did a lot of kind of thinking through how we would actually track this towards revenue, um, recognizing the importance of it, um, especially for the marketing personas, kinda the first try with this and, um, and really let the content run from there.

[00:15:33] Emily Sandison: So it took multiple iterations, I would say, of how forward do we wanna be with this. I think at first we were a little more like sheepish about it, of like mm-hmm. Oh, and eventually we got to the point where I think the headline on the page is literally Grammarly help write this page. Mm-hmm. Which was really cool to see even, you know, amongst the team as we just kind of came to this evolution of like, why are we hiding behind it?

[00:15:55] Emily Sandison: Like it is a really integral tool. 

[00:15:57] Emir Atli: Yeah. Um, 

[00:15:58] Emily Sandison: in our toolbox. So, [00:16:00] um, yeah, so there's the main page there and then that actually leads to some follow up nurtures, um, where we've developed what we're calling blueprints and they're really blueprints for, um, brand guides and um, snippets, which are two features that I think are really key for the enterprise so that as you use AI, your brand guides are still popping up all over the place, so folks are still writing within, um, the Grammarly tone or you know, the company tone. Um, and then some of the snippets there that just kind of help, uh, move things along faster. Um, and so there's a couple nurture paths from that, that content.

[00:16:33] Emily Sandison: Um, thus far it's been super well received. Um. So we'll see. We'll see how the numbers work out in a couple months. Um, but it's been a really fun campaign to, to bring to light. 

[00:16:43] Emir Atli: Yeah. Were you more reserved about showing Grammarly wrote this or like, AI wrote this because there's like a negative connotation, but with that, like, does AI content I'm not gonna read it kind of thing?

[00:16:53] Emily Sandison: I think so. Yeah. Yeah. You know, there's been so much on. Everywhere of like, oh, if you see an M dash M dash, yeah. It's [00:17:00] like, you, we don't wanna fall into that, um, that camp. Um, and so how do you do it thoughtfully where you're like, it's, you know, I am using this. It's really important to use AI, but it's also like thoughtful and with intention.

[00:17:12] Emily Sandison: Um, and I, I think we struck that balance. 

[00:17:16] Emir Atli: Yeah. I think it, I mean. I feel like it's shifting a little bit more and more. It, it was, it was like M dash, it's AI, I'm not gonna read it. Now it's more like, oh, there's actually, there's a good LinkedIn post. It's a good blog post. Or actually there's some AI help in this.

[00:17:31] Emir Atli: Actually, I wanna learn more about this. Like, how did they do it? So I think it's shifting over time. It's probably, it would take probably more like six to twelve months to basically be fine. Totally fine with it, but I think it's shifting over time. 

[00:17:44] Emily Sandison: Yeah. Yeah. I mean, are you writing your, like all your LinkedIn posts from scratch, are you.

[00:17:49] Emily Sandison: Combo ride? Like what's, what's kind of your take on it today? 

[00:17:53] Emir Atli: Yeah. Um, for my Linked, so I, I post, uh, every single day on LinkedIn. I mean, [00:18:00] sometimes I miss days, but it's 90% every day. Right now, how I use it is essentially I have like ideas myself, and then I have like a GPT. GPT is trained to not. Write the entire post.

[00:18:12] Emir Atli: So no matter what I say, it's not writing the entire post. It's mostly giving like bullet points or ideas from like other posts that I wrote or the other posts that were successful, my profile and stuff like that. So I use it mostly for like metaphors. Like for example, I write something and I want to do a metaphor.

[00:18:29] Emir Atli: I cannot think about it. So it writes the metaphor for me. It gives alternatives. Um, what I've seen success with is. If you ask AI to give alternatives versus like writing something, it makes you more creative. So I don't, I never use the end output of Chat GPT. I use Chat GPT mostly. I usually ask to bring me three, four alternatives and then I mix them and then do something on my own, which is generally more creative than I can be on my own. Yeah. That's how I use it. Yeah. So for example, to give you an example, if I have like a post, I don't know, like we do lots of commercials. If I [00:19:00] do a commercial, I basically write like this is the script of the commercial. Um, I wanna write a post that would uh, um, I don't know, it's like a lazy CRO commercial.

[00:19:08] Emir Atli: I wanna do like a day in my life kind of post. What would be a funny way, gimme like a couple alternatives and it gives like lots of crazy ideas and creative ideas to do like a schedule on that post, and I use it in my post, but it's overall my idea. Yeah. That's how I use it. 

[00:19:23] Emily Sandison: Yeah. Yeah. And I think, yeah, I think that's where the nuance is, right?

[00:19:28] Emily Sandison: Where it's like it can totally augment and help, um, coming from blank, you know, blank page probably. Mm-hmm. Yeah. Junk in, junk out a little bit. 

[00:19:39] Emir Atli: How is all of these changes and shifts in the market impacting your hiring? Or like, I'm sure you're interviewing people constantly or like you're thinking about hiring.

[00:19:47] Emir Atli: Is there any changes on that side? 

[00:19:49] Emily Sandison: Yeah, I mean, I think like we're seeing in a lot of the market, like really taking a look and saying, you know, what can we do with the team? We have, what [00:20:00] can we do with the tools that we have, um, before jumping straight into hiring? Um, and so, um. We have critical roles we're still hiring with, but also adding in that layer of folks that, that we are hiring and saying, you know, where is kind of that AI proficiency in there and, and making sure that that's a key part of how people are starting to, to work.

[00:20:20] Emily Sandison: Um, I think regardless of role. Mm-hmm. And so, um, you know, I don't think we've seen like huge changes on that front. Um. But I'm so curious what the next six to twelve months look like, um, in the space, you know, for tech companies in, in general. 

[00:20:37] Emir Atli: Yeah. Yeah. I think, um, maybe the hiring or like I. Uh, hiring might not be changing, but I think the interviews are changing a little bit more towards like, things that you cannot do with chat GPT or questions that might take more like reasoning on your own or like, I don't know, take on projects and things like that.

[00:20:56] Emir Atli: Or like even working together for like a day or so. I see [00:21:00] more and more of those and I, I try to do those a lot for our, especially for our critical roles to see like how we work together, how we would work together outside of AI, I mean. Everyone should use AI, but I would love to see how we would do like a white boarding session, how we would think together, I would reason together, like how we would solve problems.

[00:21:17] Emir Atli: I think those are becoming more and more important in the hiring process. Yeah. Especially for us. 

[00:21:20] Emily Sandison: Yeah. No, I, and I think that kind of just gets to that, you know, even the similarities with field marketing, right? It's that human element. Yeah. To say like, we all know it can scale and move and work in a virtual manner, but what does it look like when we sit across the table from each other, because.

[00:21:36] Emily Sandison: You know, if there's not any sort of jiving there mm-hmm. It's all gonna fall apart. Um, and so, yeah. You know, I think it's a little bit of both. Um, and I am curious as, as AI grows and just becomes, you know, for the majority of people, a key part of their work, um, how that human piece is actually maybe gonna become more important.

[00:21:59] Emir Atli: Yeah. [00:22:00] Um. And are there any, like, I don't know, I'm sure you get slacks all the time. If like, oh, there's a new tool. Should we use this tool kind of thing is, are there any tools that you would like to test right now or you're looking to use? 

[00:22:13] Emily Sandison: Yeah, I mean, on the side I've used, um, Lovable. 

[00:22:20] Emir Atli: Mm-hmm. 

[00:22:21] Emily Sandison: Just for fun websites and yeah, like that's been actually pretty astounding, just for like personal stuff to be like, oh, I'd like to create a website about, you know, this, that, or the other thing, and how quickly you can do that.

[00:22:31] Emily Sandison: Um, so that's been a kind of interesting, just on the side, I think watching what Clay is doing is really interesting. Um, and that's a tool that I'm just like, whoa, they're, they're moving really, really quickly. Mm-hmm. Um, so I've been watching kind of both of those, those players. Um. Just, you know, personally for how that works.

[00:22:52] Emily Sandison: Um, I do watch you guys quite a bit as well, and like, um, congrats on the growth. It's, it's really cool to see from a marketing perspective [00:23:00] everything that kind of goes together there. So, um, yeah, I, you know, there's just tools coming left, the time and center. Yeah. Um, I have, um, a friend who's dyslexic and has been really cool to see, you know, what he's been posting, um, and, and tools that are like really helping on that side of things. So I think, you know, there's a whole suites of tools on the enterprise side and all of these tools we're finding on the sides, you know, for our own, um, learning and own development, which kind of all come together. 

[00:23:31] Emir Atli: Yeah. Well, um, before we wrap up, I would love to learn about like, what are you excited about in the next six to twelve months?

[00:23:38] Emir Atli: What are things, I mean, it might be things that we talked about already, but in general is a demand, generally there's a marketing leader. What are you excited about? 

[00:23:45] Emily Sandison: Yeah. I am, I'm, I mean, I'm personally excited in that, um, it, it feels like there's a turning point in that the, the average person, the novice person, can totally use [00:24:00] AI effectively.

[00:24:01] Emily Sandison: Um, which feels like a really big turning point. I feel like, you know, three to six month ago everyone was like, are you, you have to be a prompt engineer. Um, and that's not the case anymore, right? Yeah. Like, you can get really, really good outcomes. Um. Without having to have these like, masterful prompts.

[00:24:17] Emily Sandison: Yeah. And that makes me really excited because, um, you know, then we can all use it and it's not just for folks who can code or things like that. Like it's, it's democratized. Um, so I'm really excited about that. I am excited about, um, just being able to kind of create the personalization and tone. Mm-hmm.

[00:24:36] Emily Sandison: Um, I've seen just personally like a pretty big jump there in the tools I've been using, um, where I'm like, oh, this actually. You know, sounds like me. Um, and doesn't need like seven different rewrites. And so there is a personalization there that I think is coming along, um, that will get there. Um, but more so than anything, just how much more quickly it helps synthesize data, [00:25:00] move through, um, kind of internal marketing stuff.

[00:25:03] Emily Sandison: Um, and. I'm starting to do to use AI in this way, but I think this is only gonna grow is like, um, really asking it to be a critic mm-hmm. Of my work. Um. I think when I ask, even my most trusted colleague who I know would give me like really direct feedback, like, I'm gonna get more direct feedback over here.

[00:25:23] Emily Sandison: And so I've been doing that quite a lot. For any sort of campaign briefs or plans or anything like that, just be like, be a critic, like, really jump in here and I'll keep pushing it to be more aggressive. Um, and the feedback I've been getting like really helps kind of overcome some of the blind spots before, um, presenting it to a larger group or something like that.

[00:25:43] Emily Sandison: So, um, it's actually really helped with dependency planning too. Um, I. Was working on a kind of a Q3 plan and, um, going through the process of like, beat this up, beat this up, and it finally was like, this plan is really content heavy. Um, and so I [00:26:00] went and checked, um, with the content marketing team and was like, do you guys have enough time for this sort of thing?

[00:26:05] Emily Sandison: And so we kind of did a good like resource check there, um, which I think is a really cool, uh, way of using it. Um. Again, just relationship aside to just be like, be really harsh with me and tell me what's wrong with this. 

[00:26:16] Emir Atli: Yeah. I genuinely believed that at some point that we would all hire prompt engineers.

[00:26:22] Emir Atli: I feel like we were all seeing it on social media, like hundreds of job postings on prompt engineers. I don't know where they went now, so I never see it. Also to the last point, I. Definitely use it the same way. One interesting thing that I've found is before, so for example, if you write something with chat GPT, if you ask it to review its own work, it actually gives lots of good feedback before even you get to last output as well.

[00:26:44] Emir Atli: Yeah. It's super helpful to review work. 

[00:26:46] Emily Sandison: Yeah. Yeah. I think, um, just doing a gut checkup on most things and, and kind of saying like, this was in my head is, does this make sense? Um, it's been really helpful, so I, you know, I'm optimistic about it. Um. The [00:27:00] more, as the days go by, all of a sudden it's like, whoa, this feels more of a speed up rather than like a real slow down to learn.

[00:27:07] Emily Sandison: Yeah. Um, and, and that's pretty exciting. 

[00:27:10] Emir Atli: Definitely. Well, Emily, thank you so much for joining me. Are there any other final thoughts that you wanna share? 

[00:27:15] Emily Sandison: I don't think so. No. Thank you so much for, for having me and, um, yeah, really excited about this podcast and, um, it's been fun. 

[00:27:23] Emir Atli: Cool. Thank you very much.

[00:27:24] Emily Sandison: Yeah.

Episode Takeaways

In this episode of AI for Go-To-Market, Emir Atli sits down with Emily Sandison, who leads demand generation at Grammarly. From PLG signal chaos to humanized AI content, Emily offers a rare window into how a scaled GTM team balances internal tools, creativity, and AI adoption, all while maintaining empathy and clarity in a fast-changing environment.

This is how a demand gen leader builds strategy and trust at enterprise scale, one smart AI experiment at a time.

1. Signal Overload to Strategic Focus: Cutting Through the Noise in PLG

Grammarly went from zero to hundreds of buyer and product signals in just six months. Emily’s team is now zooming in on the three to five that actually matter—cutting through the noise to drive high-precision sales plays. Her team uses AI to summarize patterns, bucket behaviors, and make these signals usable for reps without creating paralysis.

“We’ve dumped a bunch of signals on people. But now we’re slowing down to speed up, refining down to the ones that actually drive closed-won.”

Takeaway: Don’t drown in PLG data. Instead, use AI to surface patterns and create signal playbooks that your sales team can actually act on.

2. Internal vs. External AI: Build Where It Counts, Buy Where It Scales

While Grammarly leverages ABM tools, they’ve also built their own internal AI models to make sense of product usage data. This hybrid model: buy externally, build internally is helping the team move fast without over-relying on noisy vendor stacks.

“There are areas where tools overlap but we let both run to learn, then consolidate. That’s how we find what actually works.”

Takeaway: Use internal AI builds for strategic edge and flexibility. Use vendors to accelerate experimentation and be ready to trim what doesn’t compound.

3. Content, But Make It Human: AI as a Creative Amplifier (Not a Replacement)

Emily is clear: AI doesn’t replace marketers—it enhances them. Grammarly’s Customer Zero campaign was built entirely using Grammarly’s own tools, but every word still had human review. The result? A marketing campaign that proudly declared “Grammarly helped write this”, without sounding robotic.

“We were sheepish at first. Now the headline is literally: Grammarly helped write this page.”

Takeaway: Use AI to co-create content that feels human, contextual, and brand-safe. Let it spark creativity but never outsource judgment.

4. AI Skills Will Be Expected But Human Judgment Is Still the Differentiator

Hiring at Grammarly now includes looking for AI literacy across roles, but emotional intelligence, adaptability, and curiosity remain essential. Emily is especially excited about using AI as a “critical feedback loop”, asking it to poke holes in strategy decks and catch blind spots before going live.

I ask AI to be a critic. And honestly? It gives me direct feedback and helps me overcome blind spots.

Takeaway: AI fluency is table stakes. The differentiator is using it with intention to speed up learning, reduce bias, and deliver sharper campaigns.