Tim Harrison

Great to have you here, Varun.

Varun Anand

Yeah. Thank you for having me. I just wanted to give a big shout out to your whole team. You know, Chris Ross, Jack and your team, they've been really great to work with. And I feel like Hg has been, you know, on the cutting edge of a lot of this AI stuff in private equity and really being at the forefront of what AI can do on the application layer.

And it's been great to work with your team. I love that summit we did for all your CROs in Vienna. And yeah, I'm excited to do way more with you guys.

Tim Harrison

Likewise, the partnership with Clay has been incredible, and not only in the Vienna Hackathon where all the go-to-market leaders were together hands on, but also hearing the CEOs talk about it in the Valley this week.

You've had a very unique journey that brings you to Clay. Could you tell us a bit about that journey and what brought you to the role at Clay?

Varun Anand

Yeah, well, I actually started my career in politics and government, so I worked for Hillary Clinton for a number of years. And ultimately, you may remember, we lost a certain election in 2016. And that forced a pivot into technology.

And I was at this insurance company where I was the GM of one of our business units, and our CEO wouldn't give me any engineers to work with. But he was like, you still got to meet these revenue targets. So I actually built a lot of product using no code tooling like Bubble and Excel and things like that. And it worked. But the bar in insurance is very low. So it's like not so hard to win with things like that. But that's kind of how I got into this world.

And Karim actually had started Clay originally in 2017. But, you know, they were exploring a lot of ideas. They hadn't really committed to a particular product direction. And I left to start a company, and I was in, like, you know, eight different Slack groups about no code tooling because I just spent so much time on it. And someone posted about a webinar for Clay. But at the time, they didn't have any customers or revenue. And so I show up to this webinar and I'm the only one there.

No one else is there. And they were going to cancel it. But I'm like, okay, this is interesting. And I started using the product in this early prototype and I was like, whoa, this is interesting. You know, like, this is cool. And so I cold-emailed Karim, who doesn't respond because he's on a silent meditation retreat. But his co-founder at the time, Nikolai, responds, we hit it off. Start spending more time together. Start working together. But then Nikolai leaves. Everyone else also kind of leaves.

And for whatever reason, Karim and I decided let's work together. And, you know, I kind of join as a late co-founder. And then we kind of reset the business, focus on go-to-market as our core use case, and launch the product and start growing from there.

Tim Harrison

And that refocus, that sort of resetting of the vision - how did that come about? Was it very clear that go to market was going to be the direction?

Varun Anand

I think it was clear, you know, I think there was a lot of pull that we were seeing. Now, where that pull was specifically, I think was an open question because go-to-market is a massive space.

But I think we always knew - and I think credit to Nikolai, who had this insight originally - that salespeople in particular have this real kind of manual work that they're doing, always in spreadsheets. But we actually started with cold email marketing agencies, like we started with these really small customers. And it makes sense because they actually - they're very technical, they're scrappy, they're entrepreneurial, they feel the pain most acutely because they have lots of clients, and they were able to use our product first, and then they started posting about us on LinkedIn.

And then we got a lot of traction through that and then expanded to their customers, expanded to startups, went to the biggest enterprises in the world now, including Salesforce, OpenAI and Nvidia, Google. And that's kind of how the initial traction started.

Tim Harrison

And then where did the launch of Gen AI and Gen AI into the…

Varun Anand

Yeah. So we were already getting traction. We, you know, for the first year and a half. And then the ChatGPT API came out. Well, ChatGPT came out itself, then the API came out. And that basically accelerated everything.

And I think it did a few things, right? Number one, it actually completed the workflow because Clay was no longer a data enrichment product. It could actually do more than that. You could write and draft emails. Number two, you could use AI in Clay for like data analysis, data scraping, like data cleaning, all that type of stuff.

You could use AI in Clay in generic ways to go out to websites, take actions, get information, bring it back to the spreadsheet. And you could use AI in the product to make the product easier. We had JavaScript that you had to write in the product in order to do certain manipulations, and now you could just write in natural language to AI, and it would translate that into code and actually do work for you.

So a lot of that really accelerated the business and helped in a lot of ways. And it was - the form factor of an un-opinionated spreadsheet and workflow builder was the perfect kind of setup for AI to really supercharge.

Tim Harrison

That theme of looking at the full end-to-end workflows rather than just the specific pieces, is one we see across all the different functions when looking at AI use cases. So it's really interesting how you see AI actually unlocking those end-to-end processes.

Varun Anand

No, no, for sure. And I think, like, you know, Clay, basically what we do is we help go to market teams find their best customers, find more like them, expand them, right? And that end-to-end workflow like you're saying is comprehensive.

So it starts actually with the data. This is not like the sexiest topic in the world. But let's be real. Every B2B company has shitty data. Like they all do. And they're usually spending more money than they like on it, have more vendors than they want on it, and it's at the end of the day, shitty. Let's just fix that. Let's consolidate the vendors, let's dramatically increase the data quality, and let's give you a foundation on which AI can actually be useful, right? And that's where we start.

And the way we do that, by the way, is we aggregate all these vendors. So we have all the vendors in the world. And then we give you AI on top to find the specific data points that you need. And then once you have that, then let's do the full workflow. And that could mean, hey, let's do account-based marketing, let's send direct mail campaigns to people, let's do automated outbound at scale, right?

Let me give you a very clever, fun example. So one of our customers - the trash company we all know and love. And they use Clay to look at Google Street View images of people's homes. Then they use AI to analyse that, and they look at the colour of the dumpster outside your home. And then they're like, okay, well, what colour is it? Which indicates what competitor they're using. And depending on the colour they run, they use that to create an automated, targeted direct mail campaign to convert the homes from their competitors to Waste Management. That’s a pretty end-to-end workflow.

Tim Harrison

That's incredible. Just reflecting, when I go back five, ten years working on data analytics - called data science back then - projects for enrichment for sales. It only took a few pieces to go wrong before you lost a lot of trust. How have you got over the trust and data quality?

Varun Anand

The accuracy is better. Okay. Like when we started working with OpenAI, their data quality is 40%, so obviously you can't trust it. Now it's like north of 85%, you know. And that was just immediate. With Anthropic we started at 30%. Now they're north of 90%. So when the rates are that high, you obviously are going to trust it way more, and you're obviously going to have way more faith in that. But if you're only right three out of every ten times, of course you're not going to trust it.

Tim Harrison

And one other theme I've sort of reflected on across the different functions, particularly sales, is there's one part which is buying the new tool or buying the new data set. And then there's the actual operationalisation of those insights and how you rethink about those processes. And with Clay using AI, right, that's quite fundamental reimagining. How have you seen the most successful teams using Clay actually sort of operationalise and reorg and transform around how they use it?

Varun Anand

Yeah. Well, I think just as a broader point to your shift on AI, the first thing is, as things like Claude Code and stuff like that get more popular, like as these coding tools get more prominence, software gets commoditised, right? It's easier and easier to build software. It's easier and easier to vibe code things. And so the real differentiation is in go-to-market. It's in how are you getting to market. It's how are you actually finding your customer? That's where the edge is, right?

And so that's why you have to change your orgs and why you have to adapt in order to win right now, because everyone is doing this. And so the bar gets higher and higher, right?

And so what we're seeing actually is the emergence of this role that we invented actually called ‘Go-to-Market Engineering’. And we're seeing, you know, the best companies in the world - Cursor and Webflow and OpenAI and all these companies - hire Go-to-Market Engineers and centre their operations and work around them.

And these are the people who are building revenue systems using AI and automation. And they are having ideas, testing experiments, running things and playbooks. And they are actually trying to grow revenue. And they are the ones coming up with these plays and using Clay and other products to execute them.

And they're operating more like a growth function where like data science supports it, than like an ops team that's just managing a system.

Tim Harrison

And I'd love to double click on that. We're having some of our CEOs this week. They're talking and asking about this role of the Go-to-Market Engineer. Is that something you sort of live and breathe at Clay yourselves? Do you have that role?

Varun Anand

Yeah, yeah. Of course. I mean, we invented the role. We have the role. And by the way, we not only have the role in the traditional sense, but we also have them sell Clay. So we also have Go-to-Market Engineers be our sellers.

Because, well, because we are - like, who better? It's like if you were selling design software, do you want sellers selling it or do you want designers selling it? Like who is going to have the empathy? Now, designers are not always the most charismatic people in the world. So maybe you don't want them selling your product. But in this case, go-to-market engineers are charismatic and they have way higher empathy, way higher trust because they are in your seat and they are the one who know what you're trying to do and have empathy for that and can help you get there.

Tim Harrison

So one question that we've been reflecting on, if we look back at 2025 and 2024, is AI in support and AI in R&D have been easier, probably nuts to crack, and especially when it comes to that transformation operationalisation piece. We're starting to see really successful pilots. And you'll be speaking and talking to some of our portfolio companies this week. What do you think's made it harder about go-to-market? Or is that actually you're seeing differently across…

Varun Anand

No, no. Yeah. I think actually even today I think there was an announcement - Anthropic this morning, I think there was an announcement that shook the public markets a bit that Anthropic is moving into the legal space, you know. And I think there's reasons for this.

Basically, I think it's obvious what LLMs can do. They can translate text into other things, right? And what is that really useful for? It's useful in support. It's useful in coding. It's useful in legal. These are very straightforward, obvious applications of AI.

The difference though is go-to-market is not like that. I think I have a comprehensive answer here, but I'm going to go into a little bit of a history lecture. So bear with me for a moment.

Basically over time there's only been two outcomes in go-to-market that have really been exceptional. That's Salesforce and HubSpot. And isn't that weird? This is the biggest market. It's sales and marketing spend. People are spending hundreds of billions, trillions of dollars on this and only two companies have emerged. Isn't that strange?

And both of those companies, by the way, are systems of record. They're not even like applications. They're systems of record. So what is up with that? This is actually, I think, the answer to your question.

The reason why I think that is - is in go-to-market, things change all the time. You're an investor. You work at Hg.

You're an investment firm. Your entire business relies on what you call alpha. You know, your ability to spot something different or unique or interesting about a business and have that unique insight that gives you some edge, that you capitalise on, that leads to your overall performance in the markets, in that investment.

Go-to-market is the same thing. It relies on alpha. And so like basically you go to market, there's all these products that come up and they grow quickly because it's just like, hey, do you want more money? And they're like, yeah, obviously. So then you pay for that. But then it works, right? But then it stops working because things change in go-to-market. Because as you create an edge, as you find alpha, someone else figures that out and they copy it, right? And so the edges don't last.

Now AI has emerged, which means these edges don't last as long as they did before because it amplifies everything. AI helps everyone do everything faster. And so if you find some edge, someone else is going to find it. It may have taken three months. Now they find it in two weeks.

So what do you actually need? What you actually need is something that helps you iterate faster - a system, an engine that helps you experiment, to iterate more quickly. That's what you actually need. And that's why no product has broken out. Because all these products are very opinionated. They're built for sales reps. Because you know why they're built for sales reps? Because they're all based on per-seat pricing models. And you know who has a lot of seats? Sales reps.

And so if you're built for sales reps, you build a product for everyone. Because sales reps are the common denominator when you're building a product. And then you don't have an edge, because if everyone has access to the same information, there's no edge.

So we've actually chosen to build for the more technical operator - the RevOps person, the Growth person, the Go-to-Market Engineer. And we have built an un-opinionated primitives-based product that gives them the option of taking those primitives and applying it to their business in a unique and tailored way.

And that's why even as the trends change, they will be able to adapt and come up with new ideas and execute them in Clay. And so that is a very long-winded answer as to why this has happened in go-to-market and why coding has made sense, why support has made sense and go-to-market has not taken off yet in the same way.

So coming back to your question - that's why in coding and support, AI has gotten so much prominence so quickly because it's such an obvious, straightforward, top down application of LLMs in this context.

But go-to-market is different. Go-to-market actually relies on alpha, which we just talked about. It relies on your insight into your business, into your customer. It relies on creativity to come up with that insight and act on that edge, and then have another insight when people copy it. And that is something that AI can help with, obviously, but it's very different and very unique and keeps changing and relies on insight and human creativity.

And that's why I think it requires a very different approach in go to market than it has in coding or support.

Tim Harrison

That's fascinating. You mentioned there the creativity and innovation side of it. You will have seen across many companies using Clay - have you started to see any traits or attributes of the culture of the teams and even the employees and the people where Clay gets adopted and drives the most value versus ones where you hit friction?

Varun Anand

It's teams driven by curiosity. You know, it's teams that are technical, that are scrappy, that are resourceful and that are curious. And curiosity comes from being able to use a product like Clay and use it well and use it to its fullest potential. But it also comes into curiosity into your customer base, and having a deep understanding and insight into what's happening, into what resonates with them and really relying on that.

Tim Harrison

And what patterns do you see in those go-to-market teams that actually translate that adoption and that curiosity into performance gains and win rates? And how do they productionise and scale that?

Varun Anand

Yeah. So I think first of all it's investment, right? So you need people to do this and you need the right people to do it. And you need to set up a team. And maybe it's a go-to-market engineering team that's actually focused on this. And you're giving them tools. And sometimes they're products like Clay and sometimes you're building tools internally. It doesn't have to be like third party software. And then you're building this culture of iteration and experimentation.

That's kind of what the best teams are doing. And then they are using it in like a very systematic way. And so you're basically building an engine, a system of intelligence where you're always looking for information and getting it right.

So for example, let me tell you a quick example of how Clay uses Clay and has built a system like this. So for example, let's talk about social listening. So we built a social listening engine. And this engine using Clay is looking at my LinkedIn posts. It's looking at anyone who mentions Clay. It's looking at our customers. And it's saying okay what's happening. So it's doing sentiment analysis on what's happening.

And if there's a negative sentiment, we act on that. If there's a positive sentiment we route that in the right place. If it's a sentiment from a customer, an existing customer, we route that into Salesforce. We include that in an account briefing for the rep or the CSM, so that when they meet with the customer next, they have that context, right? So it's always happening.

If it's one of my posts and we use Clay to analyse the people who commented on it, who liked it, and to understand who are the ones who are prospects, and then we get that information so that we can reach out to them. And that's always happening all the time, right? And I don't have to think about it. And so you've built this very unique system using Clay.

Another example would be competitive mentions, right? So we built a very unique system with Clay that is always scraping, you know, Gong calls. That's looking on LinkedIn to be like, okay, how are we talking about Clay? And then once we get that information about competitors, we can take that to the product marketing team. We can feed that and give it to the reps during their account briefings for a relevant meeting, right? We can take that and we can put it in the CRM and update it in the right place. We can take that information from a Gong call and then update the next best action for a rep.

And this is all happening systematically without someone having to think about it. And that's a very different way of operating.

Tim Harrison

And have you found that sort of innovating and using your own product is something you can then build on and take to the product?

Varun Anand

Yes, yes. And that's actually a very tight loop. We have PMs in the company whose entire job is just that - like our go-to-market engineering team internally, their job and mandate is to push the boundaries of what's possible and take those. And they don't have to use Clay to push those boundaries, by the way. If Clay can't do it, do something else. And then that creates the pressure for the team internally to do it.

Tim Harrison

And then sometimes a harder question, which is where have you seen it not work or where have you seen resistance from sales teams or where has Clay not been able to adopt?

Varun Anand

I think it comes down to investment. So it comes down to like, hey, do you have a team internally that can own this?

And then the other pushback is change management, right? Hey, we might have a thousand sellers and there's fear around that. Now, of course, we don't want Clay to be used in its present form by sellers. So we're building a system that is meeting sellers where they are, right? Maybe they're getting DMs on Slack. Maybe they're getting text messages. Maybe they get the information so they don't have to go to Salesforce and update. Our reps aren't even touching Salesforce anymore, right? They're getting the information where they are, and they don't have to log into Salesforce to update something. That's happening in the back end through Clay.

But there's a lot of fear around change management, and there's also a fear if you haven't made that investment into a centralised ops team, a go-to-market engineering team to make all these changes.

Tim Harrison

And I think that change management piece and the need for the real transformational grit and hard yards to transform and reimagine what the team looks like and how it can reorg around these new processes.

Varun Anand

And I think we've seen that even within your portfolio. I think we're proud to be working with more than half of the portfolio already and hopefully the full portfolio by the end of the year.

And, you know, I'm doing a talk with Scott at A-Lign later today and I think they've had great success. Their data operations work was taking 48 hours, and now it's less than two hours, right? So it's like 24 hours faster. And so what Clay is able to do is increase the top line. Right? That's higher pipeline. That's higher revenue. That's greater quotas. But it's also reducing the bottom line. So that's saving you time. That's taking BDR manual efforts from like five hours of manual research a day to less than one hour.

And that's kind of the - you get it on both sides.

Tim Harrison

And it's time they can be spending on higher value add tasks, right?

Varun Anand

Having real conversations with prospects. And actually being way more creative in how they engage with prospects. Our sellers are building vibe coded apps in Lovable to engage prospects more creatively. And they're doing that in much more thoughtful ways because they don't have to do manual work.

Tim Harrison

And do you see it actually giving a better experience to the end customer that's being sold to?

Varun Anand

Yeah, I think because ultimately customer experiences are driven by how empathetic you are or how much you understand them, how tailored your sales process is to them, right? And if you are not doing manual work like updating Salesforce or manual research or whatever, you have time to invest in making the customer experience as thoughtful as possible.

Tim Harrison

The team actually let me know is that there's no product managers at Clay?

Varun Anand

We do have product managers. We do not have - we do now have product managers.

Tim Harrison

It's interesting. I'd love to talk about that journey or is that less relevant for you guys?

Varun Anand

I think it's just that for a long time we didn't. Yeah. Now we do. We don't have that many. We have like four, you know, so it's not that many PMs. But it's just that, especially when you're early, you want really tight feedback loops in the business, right?

So that's why for a long time our support was done by engineers because you want really tight feedback loops. The people who are helping our customers are the people who are building our features, right? The same thing applies in product. You want feedback loops to live within one person. So you don't want like a PM and a designer and a this and that, right? You want the engineers to understand it themselves. And so they are living in that feedback loop.

Tim Harrison

That's a really interesting piece we see, especially when looking at AI and R&D, is that role between the product manager and the engineer almost getting blurrier? Or having engineers - what you call the forward deployed engineers or engineers closer to the customer - but also product managers who are more able to develop and capture that engineering skill set. Just like the Lovable example you talked about.

Varun Anand

Yeah, exactly.

Tim Harrison

And then one sort of forward-looking note to end on. So much is changing in AI and go-to-market at the moment. Where do you see AI and go-to-market in - let's choose a horizon of 18 to 24 months?

Varun Anand

I think it's a very exciting time. I think basically the big push from now until then will be what can actually be done autonomously, right? What agents can actually be working on accounts for you? How can they be growing your accounts for you? How can you go from these one-off, 1-to-1 actions to one-to-many actions, and have entire systems that are working on your behalf while you're just, you know, focused on the highest value human-led interventions?

And I think that's kind of the biggest leap that we have over the next couple of years.

Tim Harrison

And how are you seeing the go-to-market leaders drive that most in their teams in terms of their culture?

Varun Anand

Well, I think a lot of go to market leaders that really believe this and feel this are leading by themselves, right? And so, you know, just for example, Hg and Clay partnered on a hackathon, in Vienna actually, with all of the CROs from the portfolio. And we sent several people from Clay and they were very hands on, you know, knees deep in the product, building things, creating workflows, making an impact on their business.

And so they get it. And once you get it, once you understand it, you can create that pressure and you can create that culture in your team. But it starts top down, obviously.

Tim Harrison

And we've seen that across lots of different functions in AI. And also for the transformation, it's making sure there's that exposure. And the leaders are in the boiler room and seeing how this technology works and what it really looks like in anger. It's such a powerful point for them expanding it to the team.

Varun Anand

Yeah, absolutely.

Tim Harrison

Varun, thank you so much for your time. It's been great to have you on the podcast.

Varun Anand

Yeah. Thank you so much. This is great.

The views and opinions expressed in this podcast and transcript are those of the contributor and should not be taken to represent the views or positions of Hg or its affiliates.

Statements contained in this podcast and transcript are based on current expectations or estimates and are subject to a number of risks and uncertainties. Actual results, performance, prospects or opportunities could differ materially from those expressed in or implied by these statements and you should not place any undue reliance on these statements.