E49: All about AI and GTM at Clay - Round Two with Everett Berry

Hello and welcome to the Revenue Leadership Podcast by Topline. I'm your host, Kyle Norton, CRO at Owner.com. And every Wednesday we dive deep into the strategies and tactics being used to drive success by the best revenue leaders in the world. So join me as I sit down with revenue operators to discuss actionable frameworks that you can implement in your business today with no fluff, no sales pitches, and no platitudes.

If you're measured on pipeline, win rates, and forecast accuracy, listen up. When your product doesn't show up until the second or third sales call, you lose momentum and deals. Walnut fixes that. It gives your reps interactive demos they can personalize and use early, so every buyer sees a consistent story every time. The result? Faster cycles, higher conversion from demo to op, and better forecasting based on real product engagement. Measurement, not guesswork. You ramp reps faster and close more without adding SE headcount. Go to walnut.io and hit the Get Started button to see how you can turn revenue targets into reality with the team you already have.

Today's guest is Everett Berry, who's the head of GTM engineering at Clay. Everett joined us a couple months ago on a very well-received episode, and so we are doing a slightly different conversation today. This is an AI Q&A. So I sourced questions from a bunch of listeners who are the folks that send me messages on LinkedIn and text me about stuff and collected what they would want to know in a more Q&A style conversation. So we're going to bounce through a bunch of different topics today. So slightly different format than how I typically manage the docket where We take one topic, we go super deep, we try to unpack it from the very strategic down to the very tactical. And today we're gonna bounce around a bunch and just cover the stuff that you guys wanted to know more about within AI. And interestingly, as I was just saying to Everett before we recorded, a lot of people wanna know how Clay works.

Yeah.

And the, the inner operations of how Clay uses AI and how Clay manages these things. So I think we'll probably spend a good chunk of time on some of those questions.

Amazing. Yeah, it's awesome to be back for part 2, Kyle. Glad the listeners liked the, the first one and I'm super happy to open the kimono on, on Clay. I love talking about some of the stuff that we've got going internally.

Cool. Yeah, I was joking with a friend. I'm becoming a bit of a one-trick pony these days cuz all anybody wants to talk about is AI in go-to-market. But yeah, a friend of mine said, you know, if there's one trick you're known for, that's probably a good trick right now.

So yeah, yeah, no, I think you could, you could do worse on that one.

Yeah. So, all right. So let's start with some of the Clay stuff. Uh, people wanted to know like, how are the teams structured? So like, how big is the RevOps team at Clay? How many GTM engineers do you have? Where are these people situated? And, and then I'll dig in from there.

Yeah. So there are two types of GTM engineers at Clay. One type is the team that I run, which is the forward deployed GTM engineer. So we are folks that Actually, many of us were in RevOps or growth or, or marketing before we joined, but mostly what we do is we help Clay's customers evaluate and implement and scope and scope Clay, and we do a lot of building there, but it's, it's mostly external. And then we also have an internal team that's currently about 6 GTM engineers that does all of our internal ops. And, uh, they are a, they serve my team and our CX team and our, um, marketing and growth teams. And I'm trying to think if there's another another team that they— Yes, in partnerships. Thank you. And so they serve all of the GTM motions at Clay and they're kind of like a central, um, a central team that, that works on each of those. And they do a lot of, they have kind of longer term projects, but a lot of what they do as well is they get tickets submitted to them from across Clay, very similar actually to how an engineering team works. They triage those tickets and bundle them into twice a month releases that they post release notes on. And, um, and then they maintain Yeah, I could talk a lot about this team, but they—

People wanna know, so let's spend the time on this. We can definitely unpack it 'cause this is, of all of the things people wanted to hear about, the questions were like 60% like, well, what does Clay do with this stuff?

Yeah. Amazing. So yeah, so the team is run by a guy named Osman who's incredibly talented. And actually when I am building kind of how should you do this type content for customers, I often start with talking to him and his team about, about what they're doing at Clay. So maybe just starting from the, the tech stack, it's really Clay, Snowflake, Salesforce, and Gong. And those 4 tools encompass the majority of, of what we do. I, I suppose you would also fold Slack in there because we have a, um, a kind of like our, our rep tool is, is actually Slack and we have like a Slack app that, that the reps use for um, for a lot of the work that they do.

Um, and, um, and Clay customers get access to the Slack app or this is purely internal?

Yeah, so we have this like human-in-the-loop flow that works inside Clay that you can deploy to your own Slack instance. And that's the thing that my team, the external GTM engineers at Clay, use to, for example, trigger campaigns or view signals or get pre-call research or post-call meeting notes or things like that. Post-call follow-ups. And so, and so the, the internal GTM engineering team at Clay maintains that application, and then they work on the motions of all the other GTM folks at Clay. So for example, our, um, growth strategy team, which is our post-sales team, they get all of their handoff decks generated for them, and they actually get their QBR decks now generated for them too. And so What that looks like is our GTM engineering team will take the credit consumption that's happening from Snowflake. They will take recordings from Gong calls that have occurred. They will take Salesforce data and they will bundle that and format that all into a series of information that they then send over to a Google slide deck that they publish on behalf of our growth strategy team. And then when that team shows up to a call with customers, they just open that deck and, and go from there. And so across Clay, there are like all of the motions kind of benefit from this team and they will do these sprints where they work on different things and then they will also work on all the basic stuff, data quality, um, quality of life automations. We are always looking at kind of like new tools. We're looking at like, for example, a, uh, AI sales coach right now. And, and so our, that team will participate in evaluating vendors for that and integrating them into their, into our stack. But, but yeah, that, that team, uh, really operates like a, in my opinion, engineering team where they get tickets filed for them. They have 2-week sprints that they do. They have a lot of stuff version controlled, for example. Um, and they use Clay, Snowflake, Salesforce, Slack, and Gong.

Very cool. And where do they report in the organization?

Yeah, so they roll up to actually our co-founder and head of ops, Varun.

Oh, okay.

And so they, they kind of sit alongside, uh, the like first line of leadership at Clay below the, the founders.

Interesting. And so they're like, they sit side by side to RevOps. And so, and so this team does a lot of what many RevOps orgs do. How do those two teams interact?

Yeah, RevOps at Clay actually sits halfway in that team and halfway in finance. And so the finance function at Clay actually does a lot of the forecasting and sort of like, you know, obviously the overall company forecasting, but also the forecasting down to the kind of like quarter over quarter sales side. And they also kind of set, or they work with the GTM engineering team to set pipeline numbers and to kind of like set targets across that. And then, um, the responsibility for something like a pipeline target is shared by the growth team, the marketing team, by GTM engineers, and then is like sort of like facilitated by our internal GTM engineering team. So it's a little bit of like a, a matrix structure, but anything that has to do with kind of like systems or aligning the, the numbers or so forth, that kind of is shared between our internal GTM engineering team and finance. Hmm.

And is this, do you think this is like special for Clay or could other organizations build themselves in a similar manner?

I, I think the one, the one thing that is not special to Clay, but more unique is Clay as a business is almost nearly 50/50 self-serve and sales-led. And so what that means is that obviously when you have a large self-serve business, there's really it's really like pure automation that is kind of driving that.

Mm-hmm.

And, and so the internal GTM engineering team owns the automation and plays that we run for that. And a lot of those plays actually are things that impact our ability to kind of sell into our self-serve funnel from the, from the sales-led side. And so I do think like if your organization is mostly sales-led or mostly self-serve, you actually probably would have a little bit of a different structure. It just so happens that you know, some team somewhere has to kind of do the work of unifying the, what for the sales side is like top of funnel activity. And a lot of that activity actually happens within our self-serve, our self-serve customer base. And so it's not perfect right now. Actually, a major initiative at Clay is to kind of more tightly unify from the self-serve, um, business into the sales-led business and like have more of a progression that happens there. But, um, but the reason, one of the reasons that our internal GTM ops and GTM engineering team handles kind of a wider array of responsibility is because they handle the systems that, that work on our self-serve motion as well as our sales-led motion.

And then who owns the like total growth number? Like how do you, 'cause a lot of companies struggle with when there's a self-serve number. That's oftentimes owned by a product growth team and a sales number owned by sales, you, you end up with like a lot of local maxima stuff happening.

Yeah.

And silos and whatever. So how do you guys, like, how do you guys ensure that that's not happening?

Well, today it is a little siloed. Uh, I would, I would admit. Um, and that's, that's kind of our, that's kind of what we're working on actually as a company where we're, we're trying to provide a smoother path for folks, um, to go from self-serve into enterprise. And, and, you know, Clay at its heart has is very much about like, um, you know, smaller growth teams, quite honestly. I mean, that's how I, that's how I started using Clay as a, as a kind of a team of one. And I, I don't think we have any plans. A lot of people kind of think we are like moving upmarket. It's, it's really not true. We don't have any plans for that. It's just we do have, we do serve a bunch of like enterprise customers that have kind of similar problems, oftentimes data quality problems that That Clay helps solve. And so, yeah, so today it's probably less of a problem of who owns it because the, the growth team can more or less independently work on the self-serve funnel while the GTM engineering team works on our sales-led, our sales-led funnel. Over time, I actually think responsibility for the overall motion will kind of rest with some type of like GTM leader. But today at Clay, there is not a leader under Varun that has sales, marketing, post-sales, and ops reporting into them. Whereas at like a lot of companies, this would be kind of like a CRO role basically. Um, and so that structure is probably in Clay's future. It's, it's somewhat above my pay grade to like make that, you know, make that call. But, you know, for Clay, we're somewhere around $100 million in revenue. And so, you know, as, as we get into probably a couple hundred million, that's, tends to be, at least from my observation, when more of like a CRO type figure kind of, kind of becomes involved. I could also see Clay following the like multiple CRO motion, uh, motion, which, which some companies have. Stripe is probably the most notable company that has a bunch of CROs and then they report to the main CRO.

Yeah.

But, but yeah, I think it, I think it's very dynamic. It, it kind of depends on how fast we are able to get to a place that we're happy with where the self-serve motion of Clay kind of like feeds the sales-led motion to a, to a greater degree. And right now that's, that's probably like a major priority at the company. So anyone selling into Clay that is like enriching podcasts, transcripts, there, there we come. That's like a major initiative.

There you go. What, so switching gears a little bit, what other AI tools does Clay use outside of Clay?

Yeah. So we're big fans of Dust. Dust is a tool that you can use to basically like, it's kind of like an agent that sits on top of your internal company information. So Notion, Gong, Salesforce, et cetera. The thing with Dust that is interesting is, and this is kind of true in general, I think for, for a lot of sort of AI projects is we've learned over time some of the common questions and workflows that people do with Dust, and we've begun to kind of move some of those to, to Clay, but there's still like a, a very long tail of things that come up or new things that, that are happening that Dust is pretty useful for. And In particular, my favorite feature of Dust is it has, it has your entire Slack ingested into it and it can give you kind of like references to other Slack conversations, which, which I find helpful. So it's an interesting product. It's, it's likely kind of competitive to some things that like Salesforce is working on and there's some overlap actually with some of the, the Klajent MCP stuff that can like use your internal docs. But, um, it's still not that easy to build kind of like a high quality Q&A agent that works well in Slack. And I, I give credit to the Dusk team for, for doing that. Another tool that's been super helpful has been Crosby. So Crosby is a like legal redlining service. And so they do actually all the first pass redlines for our contracts. And we have, we have a legal ops process that's built out. So in a lot of cases, you know, customers will be redlining things that Crosby can't agree to, and so it gets escalated to a human lawyer, but but actually in a lot of cases it doesn't and we will do a contract kind of only with Crosby and, and that's been really nice because it's, it's cut our response time on redlines from like a few days to like hours. Um, and so we are almost never waiting on our internal team to kind of get back to customers unless there's been some sort of escalation. And so, um, and so yeah, that, that's been, that's been great. Um, Granola is another tool that's, that's popular with a lot of the team. It's on my list of things to kind of get going in a, in a bigger way. I just got an invite to it, but, um, I know that, I know that there's kind of like a few paths to kind of get call transcripts formatted nicely into an email. Gone is one. Granola is one. Actually, some of our team uses Attention. I think we've been very, been very liberal on letting folks use the tools that they see fit for meetings because it's a very fast evolving space. The, the most important thing is, at least at Clay, is that the information makes its way into Clay, um, so that we can kind of transform it in the way that our internal GTM engineering team sees, sees fit.

So they'll have Gong and Granola going on a call?

Um, yeah, we will always have Gong going because the piping into— that's the piping. Uh, Granola will also— some of the folks will use Granola honestly more for their like internal notes that they might be using to kind of work on an account. So yeah, we've, we've left a lot of flexibility for folks to kind of like veer outside the standard path, I would say, on how they are doing things. But the standard, we've, we've tried to make the standard path for actually working on a deal, like very, uh, we've set that bar very high. So if you, if you do nothing outside the standard tools, um, you, you should have a great experience. And then, you know, it's Clay and so you should be able to experiment with stuff that, that you find interesting and, um, and stuff that you like to use.

Correct. So Dust, Crosby, Granola. Crosby's definitely an interest I think will catch a lot of people's attention because hours, hours and not days.

Yeah, they just did their Series A, so shout out to Crosby and Ryan. I, I love those guys. Um, but, um, and that's an area where, you know, um, I actually have been impressed by their ability to kind of take these very black and white things and make the LLMs like it's it's not so easy to build a process where you don't have hallucinations or things like that. And when it comes to legal docs, you have to be especially careful, but Crosby team has done a, has done a nice job there.

Oh, Alfred Lin nonetheless on that, uh, on the—

yeah. Uh, yeah. Same, same as our investor, which I think is probably how I got— I, I think I kind of came to Varun with this problem where I was like, look, we need, I'm pretty sure this is automatable. We need some kind of initial thing going here. And, and he, I think he introduced me to the Crosby founders.

So. Cool. Nice. If you were to ask the reps at Clay or your FDE GTM, do you call your, wait, do you call the internal GTMEs something different than your 4 deployed GTMEs?

We, we actually call them all GTM engineers, which is a little, a little confusing because basically they all have this, like the 2 teams could be, could interop basically in, in lots of cases. Normally what happens is. The folks that are a little more like engineer-minded, they end up in our internal team. And the folks that are a little more like me who, uh, like going on podcasts and so forth, uh, they end up being customer-facing. But a lot of folks on our customer-facing GTM engineering team could, could do the role at other companies or at Clay.

Okay, cool. Um, and so if you ask the forward-deployed reps, what is their favorite day-to-day Clay use case, what would they say?

Man, I think like content creation. Um, I think like creating Google Docs and Google Slides is really popular internally at Clay. It's more nascent among our customers, but we've automated so much of the sales process that the things that are left are really around, like, it's really this like 90th or 95th percentile problem of taking the content from calls and turning them into really nice assets that can be quickly kind of given to customers. So, you know, it's a, it's a lot of effort to go from a pretty good follow-up email to like a great follow-up email and then kind of do that consistently over a multi-month sales process. And then kind of be building this deck as you go. So I actually think we, we haven't yet put a lot of effort into, into having our ops team do this because we're still experimenting with it. But I think content creation is, is, is pretty amazing. And, um, and I would also say there are some sort of paper cuts with using like Google Slides and Clay right now that are being kind of worked on by our eng team. So, so yeah, so all of that I think has, has been really popular of late. Um, if you'd asked me this a few months ago, I would've said the like call transcript follow-up stuff, but I think we have more or less solved that, at least from the discovery call. And so yeah, the, the content creation within a sales process is, is I think super, super popular. The other thing that we're playing around with a lot right now is Clay's native sequencer.

Hmm.

You know, uh, this, the sequencer is, it's not in my opinion ready yet for like high volume outbound, but it is, there are all these just messages that you need to send to customers that are like things that are happening at a lower volume. Could be marketing, could be sales process stuff, whatever. Could be post-sales. So, um, we've been playing around a lot with the sequencer and just the fact that it's kind of built into Clay and you don't have to necessarily like do a lot of more complex setup has been, I think, super nice. I just, I'm caveating to customers right now that I still think if you were sending, you know, 50,000+ emails per month, you would want to use a, a kind of a, a fully built out sequencer. Yeah. But having the ability to end to end, you know, start from an account, do stuff, and then send, send messages in Clay via email has been, has been super nice. Hmm.

Interesting. Yeah. Some, somebody is going to enter the sales engagement, sales acceleration space in the next year in a serious way, because I think it's going to be difficult for the incumbents, sort of Salesloft, Outreach, Gong to, to keep pace with AI native tools that just have so much more capability. Clay has been one of the companies I'm like, if I was in Clay, I would be thinking about building this because it could be like a massive second act. And in, in general, like the advice I give to people when they're thinking about AI tooling is like, no new surfaces. Don't give people a new surface because they already have, you know, Salesloft and Slack and Salesforce and their email, and you might have them in Gong, like that. There's already too many panes, panes of glass. But I think like Clay could be one of the few companies where you could make the argument that they have like a legitimate shot at being a pane of glass because you could consolidate your call recorder tool plus the sales engagement tool. I think you're always going to live a little bit in Salesforce, but it's interesting to hear that you're working on that. I think that that's like a natural product extension.

Yeah, the, like everything in Clay, the native sequencer is really meant for ops teams, um, at the, at the moment, but I actually have some thoughts on rep tools that, um, I, I am Kind of, um, well, TBD, there's, there's like more, yeah.

Okay.

More to, more to come on this, but I, I actually like on what's currently available, I would give Gon some credit here. They're the new API they have for Gon Engage is really nice. It's, it's the first sequencer that allows you to actually customize every message and subject of a multi-step sequence with one API call. So what you have to do in like Outreach and Salesloft and other sequencers today is you have to create these variables that you then like inject into these message templates. Templates. That's actually not necessary with GONE's new Engage API. You can create a multi-step cadence in GONE Engage, and then you can customize, you can kind of create a custom sequence per contact, which for the GTM engineers listening to this might appreciate that that is like a new capability. And we've been, we've actually been doing this a lot for customers because GONE Engage is something that customers are, you know, when people think of vendor consolidation, like sequencer, call recorder, forecasting, and that there's kind of a whole set of tools that compete there. But yeah, credit where credit's due. Gone Engage API is super nice and I've really enjoyed kind of working, working with their team on that. And actually we use that API at Clay. So we do, um, we, we do messaging through Gone Engage for a number of use cases and we've, we've switched over to that API, which is available in Clay as an action and it's super nice. So actually I think What I am excited for is I think a tool like GON or some of these other tools could themselves just present a super nice interface that is kind of like easy to orchestrate from something like Clay or some agent that sits in the background. And that I think is like a nice, a nice feature. That said, I still am finding a lot of teams that are kind of building custom rep interfaces and custom tooling. And I think the thing is it's getting even easier to do that actually with some of the like lovable-esque tools that are out there. So, yeah. So this is another space where there's, um, where there's quite a lot of dynamism, I think. And it's a great space to build in as well. I don't think anyone has really solved the kind of like interface of the future. Rocks is a tool that I think is pretty interesting on this. Yeah. Yeah. And, uh, and they, they work well with Clay as well. Actually, we, we just announced that recently. So yeah, I'm, I'm, I'm really excited for this space. It's, it's pretty clear that kind of the, it's clear to me that the world of reps using a lot of like very purpose-built tools across multiple Chrome tabs is going away. And there will be this kind of single pane of glass thing, which I think will be some type of software that is like very easily orchestrated by agents or by Clay or, and/or in the background. And a lot of what the reps of the future are gonna see in that interface are gonna be things that are like designed for them very custom by internal ops teams. So, um, yeah, we'll see, we'll see how this, this develops. I think this is like one of the great areas to be building in right now. I, I totally agree with you.

If you are a founder building that pane of glass, in a composable manner, like call me, I will, I want to see this because I very much agree with you. I think like the, we hired a GTM AI lead and he's doing, you know, building a bunch of our cold outbound email infrastructure and internal automations. But one of the things that I have earmarked is potentially custom building our tech stack. Because, you know, Salesloft and Outreach, they're built for the email. Everything's about the email. But we sell to mom-and-pops, everybody in vertical SaaS, anybody selling into like true blue-collar SMB. Like, you know, you're, you're doing work over the phone. It's not, it's not email as the core body of work. And there's, there isn't today a tool that's built for that, you know, like, um, the even the nooks and orms of the world. Like, we actually found that when we, when we had reps parallel dialing, their decision-maker connect rates, the dial-to-connect rate dropped off so badly, and the connect-to-book rate dropped off so badly that people were booking less ops, making 3 times as many dials, booking less ops.

That's interesting. Yeah, that could be like a telephony infrastructure thing, honestly, because, um, because, you know, they kind of allocate phone numbers and so forth. Whereas if you dial from the mobile that you use to call mom, it's, um, yeah, exactly. It's actually very similar as to my understanding to like mail spam prevention system. Yeah. The other thing that I'm excited here for is I, I see kind of Teams going international earlier and a lot of selling, and this is very nascent, but, um, I've got my eye on a few tools which make things like WhatsApp selling or WeChat selling easier. And you can imagine how that's very interesting channel for some of these companies that are able to kind of go international very quickly, especially if you have tools that'll help you kind of translate messaging and, and do things like that. Yeah, I actually, for founders building, I actually am very excited about the kind of like DM space. Um, there's been some early efforts at this.

So are we.

Yeah.

This is one of the, this is one of the failures of the outreaches, saleslofts of the world is it's a terrible experience for selling through text or SMS. There's no WhatsApp support.

I mean, my, my sales interactions largely take place over iMessage. You and I scheduled this podcast over iMessage and, and that's same here now. And that's, that's really, I think at the executive level in North America, that's really what's happening. And our sales process is designed to actually find the executive buyer and get their phone number so that, um, not the, not the GTME necessarily, but like myself or Varun or, Mm-hmm. Or Becca can, can kind of text them. But yeah, that is an entirely manual situation right now. I mean, it's, it's kind of helped by Clay, but at the end of the day, I've got iMessage on my computer and I'm like adding a contact and like copying in a phone number. And, and you have to imagine that's all—

not linked to Salesforce through, through that, through those channels, right?

And it's not synced to Salesforce. Um, same thing with Slack, actually Slack DMs, I've, I did. One of our largest deals this year I did entirely over Slack actually. And, and that's not, that's not synced over.

So, which is hilarious because it's a Salesforce product.

I mean, look, um, yeah, apparently Gon is adding, um, well, maybe I, maybe I shouldn't share too much, but, but I think like, yeah, I think Slack channels are a great spot for, for increased automation. And yeah, I mean, I, I, I'm glad this is kind of like a riffing podcast, Kyle, because I could take up like probably an hour talking about what could happen with DMs here, but I think like the thing that I feel is that there is a huge amount of open space, but there's not a ton of, um, amazing product thinking, at least that I have encountered happening around it. There's a company I love in this sales tech space. Yeah, there's a company I love called sent.dm, which I think looks, uh, really awesome. It's like some infrastructure that I would actually be excited to add to Clay at some point for sending, for, for kind of doing business over WhatsApp and iMessage. And so, so yeah, so, um, I think like, uh, this whole area is super promising. And the other thing that's interesting about it is, you know, there's all these restrictions. There's, you know, it's, it's, it's, if you kind of do it in a, in a, in a gray hat way to send messaging and, and use DMs, it's, it's maybe not as good as if you're able to kind of officially be like an iMessage partner or a WhatsApp partner and, and do it the right way. And so it's the type of market where it feels like there's a lot of activation energy to kind of like do things the right way. But once you do that, you can have like a very nice position versus maybe other players, which are like installing some sort of iMessage, you know, ghost VM on your machine or whatever. So anyway, um, I'm not saying those tools won't work, but I, I have to think that like, you know, no enterprise is necessarily going to roll out like like, you know, fake iMessage for their sales team. So yeah, I think it's a very, very interesting space to kind of work within.

Yeah. I think this is the first time in like a decade where like a true new rep-facing surface could emerge and win. I haven't felt like that in a while. And you know, like I do a decent amount of venture and angel and advisory, whatever, and No one wants to invest in sales tech. Like Clay was such a, Clay was such an exception from the norm because never, no one really wanted to do deals in sales tech because, you know, Salesloft got bought for whatever it was. And it's, it was probably, it's probably still worth like 40% of that today. Similar story with Gong and Outreach, like all those companies, the, the, they hit a ceiling and the venture community, I think felt aggrieved by, you know, the challenge of building a true like decacorn scale business. But then there, you know, people look at Clay and feel like it has a very legitimate shot at becoming like a super scale business. Like Maritech is like an investor's investor and they signed off on like the market size.

So yeah. Yeah, that's what they told me. I'd never heard of Maritech before and then I was kind of, you look deeper and you're like, oh, these guys are are super legit because they're quiet.

They, they just shut up and make money as opposed to banging the drum.

Yeah, they move, they move, they move quietly, which is kind of what, kind of what you want in business. You know, the thing with Clay is like, I do think a lot of sales tools are kind of built for a specific type of sale and type of market. And I'm not, I'm not convinced Clay actually— it feels like Clay can work on the motions of a huge variety of different things. So for example, like waste management is a customer and you know, they are, I was really excited about them because they are like way outside of our sort of ICP, but yeah, but actually it ended up being like a rock solid use case. And so, you know, I've spent a lot of time, I spent my whole career in tech and tech is kind of what I know, but I'm actually excited to kind of be breaking Clay out of that a little bit and and I think the flexibility can, can hopefully, can hopefully serve us there. But you know, we'll see. It's, it's, um, many of these, many of these previous companies, you know, got off, got off to a hot start too. And I totally agree with you. There's kind of like an escape philosophy that you're looking for that it's, that's, it's not so easy to hit.

Yeah, for sure.

Okay.

Changing gears a little, the, the question was, be honest, does AI mean far fewer sales reps in the future?

You know, I'm just not so convinced. Uh, that, that it does. It feels like actually really good sales is probably one of the areas that AI might be last to fall or whatever, or I think it will actually just evolve quite a bit. Like one theory that people have about AI is it might enable folks to build actually these really small companies, like a company that, um, like Cursor that is, I think they announced they were at a a billion dollars in revenue now and they, they still, they only have in already. I thought, I thought that's what I saw or, or I know they crossed 500 million, but I think they are, I think they are keeping pace, but their, their company is really small. And so one like theory of the case is that you could see this sort of like, what, what's it called? Like not disaggregation, but basically right now in tech, you have these Mag 7 huge companies and you could see them all actually kind of like split out into these specialized teams, which are, which are much smaller and much more maybe like product knowledgeable or experts in the, in whatever thing that they're working on. And in that world, actually, it's still sales in my opinion, but a lot of the, a lot of the like actions per minute of sales of, hey, how many leads have you contacted today? A lot of that goes away and there's a lot more just like solutions-oriented expert-level selling, which is actually mostly what I do at Clay and I think is, I think is like quite exciting and fun actually. And, um, I, I do think that there is a type of seller that is like maybe doesn't know the product as well, or maybe is like, has kind of spent time at lots of different companies and they're just, they're really good at like maybe, um, executing the sales process to like a, to like a high, to grinding. Yeah. That might be something that is like a false victim to, um, to AI, but I actually think the best sales is super creative. It's super solutions-oriented. That probably there will be more of that in the future. If kind of like if AI can kind of do everything in distribution, you know, if it can do everything that's been trained on, then the stuff that is truly novel will be left to people. And I think actually a lot of the best sales folks are really creative. And then that, that's actually a core thesis of Clay, um, is the, is to lean into the creativity of go-to-market.

Mm-hmm. I think that's a smart thesis and, and I agree with you. I think there's gonna be a bifurcation. I think the best reps may be, uh, much, much more valuable than they ever were before. And you could have, you know, you could have reps who are not mega enterprise reps that are now enabled to do $3, $4, $5 million in ARR every year. And I think the price, the market rate for a really great rep, you could see rise. They're not going to capture all that value. The company will obviously capture some of it as well. But I think you could see that. And I do think you could see the lower skill grinder rep be put under put under a lot of pressure as AI gets cheaper. You've got sort of the Jevons paradox thing where it's like, oh, you know, the, the cost of a token is going to collapse and these companies aren't going to have the revenue. It's like, no, everybody, like, people just want to consume as many tokens as humanly possible. It seems like there's no limit to our appetite for intelligence. And, and you could see the same thing with sales reps, like, as as it's easier to build companies and we can do more, the need for sales reps goes up, you know, as, as they become more valuable and the LTV/CAC gets even more attractive.

Yeah.

Potentially then it gets, uh, there's more need for sales reps than, than not. I'm not sure yet though. The other thing I think that's going to be interesting is the cursors and lovables of the world. You know, it's like, oh, we've created this massive company with very few employees. That is actually only relevant in like the true AI companies. And I don't think it's necessarily replicable across, you know, vertical SaaS or selling some like B2B solution.

Yeah.

Because there's also just not enough AI engineers. Like there's not enough, you know, PhD level AI folks to go around, but but you certainly are getting a significant advantage from coding copilots where I think we'll see, you could see smaller companies with big revenue, but you might actually just see way wider companies. You didn't go multi-product until, you know, $20 million ARR or $50 million ARR. And now, you know, you see companies going very wide with their product surface area much, much earlier.

Yeah, there's no doubt that like product iteration cycles I think are dropping dramatically. Um, and if if you're not— Clay is, I think, in my biased opinion, shipped fast for a long time, but we are actually, I think, shipping even faster than we were before. Adding features to a product, I think, is like, it's almost becoming like a given. Like, you can almost imagine sort of companies that are built to build just truly custom software for like all their customers, but the economics of it makes sense because they're able to build 10 times more features with the same number of engineers as they were before. And so, you know, you can imagine this, this crazy thing happening where like, where you just have companies where they do all the software for like one of their, for their customers, like everything. Yeah. Um, and so, you know, I don't know, there's, there's, it's always surprising how these things develop, but my belief is just like, I wasn't necessarily around for when this first started, but You know, with, with the internet or with computers in general, there was like an absolute explosion in new types of jobs. And I think we just don't know what they are. We just don't know what they are right now. And we have to kind of like grope around in the dark for a while. But I expect the same thing will happen. Case in point, you know, when, when I say the creativity of sales will come, become more important, a lot of people think I'm talking about like enterprise sellers, but you could actually imagine a really strong SMB seller managing a book of business of like 1,000 accounts where a lot of the creativity of their job is just instructing agents on how to handle these different scenarios and like running plays and, and like, or thinking about the pockets of opportunity, right? Exactly.

These companies with this profile, I want to go with this messaging and deploy. Yeah.

And that, that looks different than, than making, uh, 50 calls a day, but it's not, the outcome isn't that different. It's, it's, you're sort of this, like, you, you manage this team of agents which do a lot of the legwork for you and you intervene manually in different places. And then you also, of course, probably take calls occasionally, or you have to, you know, you still have to handle reporting up to the business. I, it's hard for me to imagine this like, yeah, like agents as managers set up that some people talk about actually. Yeah. Like I don't really, it doesn't feel quite right right now. So that's, I think, a fun thing. And I think that job would be super fun as well. Um, it's, it's almost equivalent to like, yeah, it's almost equivalent to being like a drop shipper during COVID where you suddenly could drop $10 million business with like you and like 3 people. So yeah, I, I'm generally optimistic on, on all these fronts in terms of, uh, employment, I guess, in the, in the AI era.

I think one of— I can't remember if you mentioned you, if you have kids or not, but I got a 3.5 and a 5.5-year-old. And one of the things that I think about, you know, related to— because I can take, I think you can take this sales analogy that we're talking about here and apply it to like many different verticals and jobs and roles. And what all of these things require is just like so much more agency, so much more problem solving and creativity. And like, that's not at all what our education system is currently built to do, unfortunately. You know, it's still a relic of the industrial age and you're trying to basically output rule followers and compliant, like compliant workers. And, but I think where AI is going in a very macro sense is putting so much more capability and power into people's hands. And so we need high agency, creative, solutions-oriented people. Our kids are in Montessori for that reason, because it's all about following their interest and their creativity. And I'm like, I don't think we're ever going to put them in the normal school system. Will people go to university?

I don't know. Yeah, I, I really don't know either. Uh, the, you know, what happened to me in university, and by the way, I have a, I have a 2-year-old, so not quite at Montessori yet, but, um, you know, maybe starting to think about a little preschool or something. But, um, what happened to me in university was I just kind of found this peer group that I did. I kind of fell in with the computer science kids. I did hackathons with them. I had some engineering friends. We did all these interesting projects. And, um, you know, of course there was, there was coursework and stuff, but at least at the engineering school I went to at the upper levels, a lot of the coursework was, was kind of like a project that you would do all semester that you would work on that allowed you to, like, like for example, I built a Bitcoin miner and like an FPGA or something like that. Cool. Um, and we like mined like one Bitcoin. I don't think we mined an actual Bitcoin, but we did like some step of it or whatever. Right. Yeah. So, Um, and so if you extract that from like the college experience, I think that's probably the ideal. And I think that I'm really encouraged by some of the experiments that they're doing around schooling, particularly in Texas. I think there's a lot of interesting stuff going on. Um, like off school and off school. Yeah. That would like lean into that. And yeah, it just, it does seem like kind of. Coursework for the purposes of memorizing stuff or learning how to do a, a task is not that interesting anymore. Coursework for the purposes of like giving you foundations to unlock creativity and understand things at a deeper level, that, that still seems worth it. But it's another area where I think like if you're an entrepreneur, it seems very exciting. It seems like certainly people are going to almost rethink every step of the education system, um, as it stands today.

All right, I'm going to ask one, maybe two more questions, and then we'll wrap. So I was surprised that multiple people asked this question, to be honest, but it was, how do I make the case to my CEO that we need more dollars for AI initiatives?

Yeah, um, that is a, that is an interesting, interesting question.

I thought everybody was AI-pilled. I, I couldn't believe that this was a thing that multiple people said.

I do think that there's a healthy amount of skepticism, um, actually, and, and it's— there's probably more now than there was a year ago, actually, because there's been a bunch of AI stuff, a bunch of AI pilots and so forth that companies have done that, that haven't worked. I think the nice thing about at least Clay— and maybe this is not true for all AI tools— but at least Clay's, it's very much a like, show me what you've got type product. So Anyone can kind of sign up for the tool, get some credits, and like build some really cool stuff. And actually a lot of the ways that we get going with enterprise sales is some AE actually builds something really interesting or they get some like out of, they get some out-of-band result on their quarterly number, their pipeline or their, or whatever that they didn't do. One thing that I have found with executives is if you are able to kind of If they're, if they ask you for something and you're able to kind of turn around a result very quickly, they're like, what the heck? Like how, like how did you do that? And the best executives are pretty interested in like when something unexpectedly good happens in their business. And so I think my, my number one advice on, on this is probably to kind of to build more, to like try more and, and, and kind of show people things. I don't, and that's how we, that's what we do with Clay and that's how we sell Clay as well. It is very hard to win like an intellectual argument against some executive who who has been through one or two poor AI pilots and just like does not believe in the technology anymore, I guess. But I, so I think, I think showing by doing is the number one path to unlock more investment in, in AI. And then, you know, if you've got something that you think is really good and really interesting and your execs and your leadership doesn't believe in it, maybe that's not the right company for you. You know, that's, I think that's it's all the dynamism that we've discussed on this call is true of the workforce too. It's a fine time to, to, to move, I would say.

Yeah. The world is going to be even more so haves and have-nots over the last little while. From that all the way from fundraising is a flight to quality. Talent is consolidating. I think people have learned painful lessons over the last few years about, you know, the cost of being in a mediocre company, the cost of being in a company where you're the third or fourth place solution. So, yeah.

Well, sad to say, venture just, I think venture has always been like that, I would say. Um, and it just, it didn't feel like that maybe for a long time. Um, but, but, you know, there typically is like one or two winners long, long term, right? So, and, and venture wouldn't, wouldn't kind of exist without those dynamics. At least in my very, I'm not, I'm not that deep into this world, but that's my understanding. So. So yeah, like we might look back at the kind of last 6 or 8 years and say this was a pretty anomalous time actually.

Yeah. Okay. Last question, then we'll wrap. When should I hire a GTM engineer?

I, uh, would hire one as early as possible. This is, if I had to rebuild, um, the team and tech stack that I, that I built in my last company, my first hire would probably be a GTM engineer, not a, I would hire like a probably like a founding AE as well as a very skilled like sales engineer or a solutions engineer type person, or, you know, whatever. That's some, some type of expert or a former, a former customer is a good one too. And then I would hire like a Claygency or some type of outbound agency. And that team I think could probably do a lot of damage in terms, in terms of helping you find like the first couple things that are working. And then you can just kind of invest in whatever areas are are sort of working. I will say the very first task I would set the GTM engineer to is building the golden list, which is a list of accounts where if you got a magical intro to any of those accounts, you'd be super happy and all the accounts have the right people on them and they're all up to date and you can explain why those accounts are that way and so forth. I actually think not enough companies have this and it's something that If you don't have it at the beginning, it makes it even more painful as you start needing to do territory carves or whatever that, that you don't kind of know how to structure your markets, so to speak. So, um, but yeah, I would hire one as early as possible. And I think the iterative nature of GTM engineering is actually maybe most useful when you can do different things every day at the very beginning. So it's obviously useful later, but you've just got a lot more inertia typically and like tech debt to kind of engineer around versus earlier when you can just kind of do things. So yeah, I would hire one as early as possible.

Cool. All right. I'm not surprised by that answer, but I think I would have to agree if I was gonna do it all from scratch today, that that's a pretty savvy approach. Well, this was a blast. I'm actually really excited we could do this format and we'll have to make it a recurring thing 'cause I get so many of these questions and Uh, yeah, no, it's, um, it's just changing all really fast.

And, uh, you know, I thought we announced a lot of cool stuff at Sculpt. Um, but a lot of it actually is still like, you know, we're still rolling it out to customers. Right. And so some of the things that that enables, I'll probably have new takes on that and, you know, by the end of the year. And so it's, it's a great time to be building, I think, especially in, in GTM. But it's, it's a great time to be building in general, I would say. Cool.

All right. Thanks again for joining Everett.

For sure.

Thank you for listening to the Revenue Leadership Podcast. If you enjoyed it, don't forget to subscribe and you can find a link in the show notes and be sure to leave a 5-star review. Share it with your network and please join me next Wednesday for another great conversation.