A Product Market Fit Show | Startup Podcast for Founders
00:14:13 · Pablo Srugo, Daniel Saks
[00:14:13] S2: You're kind of behind the scenes. Like I might be using it or somebody, some business could be using it. They wouldn't necessarily know. [00:14:18] S0: Exactly. Yeah. Okay. We have like millions and millions of business users that may not know that we're powering it, but we are. [00:14:24] S2: Tell me about this transition period. Like how do you perceive what's happening in AI and how do you decide to leave something that's grown to be so big to start something from zero again? [00:14:36] S0: So I think in many ways, like in 2023, AI was at the infancy and it was unclear of how it would manifest at the application layer within businesses. And I thought if there's ever a time to start something again, that would be it. I think in terms of just managing that transition, just full confidence to the leadership team that was in place and Nicholas, my co-founder and CEO. And I think, you know, part of my maturity was knowing that, hey, I think the business is ready to have one CEO. I think Nick's the right guy to do it and that leaves opportunities for me to start new things. So I think that that was a transition that took many years and we were really methodical about, but very excited about what's occurred there and it allowed me to get back to the startup and entrepreneurial juices that I was really excited about. [00:15:19] S2: So you started this like six months or so after ChatGPT. What was the original idea?
00:15:19 · Pablo Srugo, Daniel Saks
[00:15:19] S2: So you started this like six months or so after ChatGPT. What was the original idea? [00:15:24] S0: So really the thesis was I had big conviction that beyond generative AI, agentic AI, which can take action, is what's going to be transformative for businesses. And we were one of the first people to talk about agentic AI. [00:15:36] S2: Yeah, that's early. I mean, that was like right after the moment. Exactly. [00:15:40] S0: And where we honed in was on sales and marketing or go-to-market because from my AppDirect experience, the most important thing a business needs is revenue. And it's very hard to predictably identify who you should reach out to and how you contact them to drive pipeline and conversion. So I wanted to start a business that really solves the biggest pain that I went through. And I partnered with Wa Gao, who is a Stanford PhD in machine learning on the co-founding team of EverString, which was a machine learning startup that was bought by ZoomInfo. And that team became one third of ZoomInfo's revenue and really transformative to their growth. Wa wanted to, similar to me, start again in the agentic era, taking go-to-market to the next level. [00:16:22] S2: AISDR was the most obvious thing to do in go-to-market around that time. Did you start there?
00:16:27 · Daniel Saks
[00:16:27] S0: Yeah, I'd say like I didn't ever love the term AISDR because I believe that it's better to enable humans than replace them. But what we did really hone in on was if you could be really good at identifying who you reach out to, drafting the messaging, and then executing across channels, then you can automate a lot of the manual work that goes into prospecting with better efficacy. And one of the unique things that we did with our AI lab is we actually built models. The first model was GTM1 Omni. And what we did is we actually, through partners, had access to a lot of transactional data from campaigns. And we trained our AI on what works and what doesn't. And we were able to prove about a 4 to 7x uplift in efficacy for outbound campaigns. And that was really the formational secret sauce. And over time, and this kind of applies to your listeners, what we saw is that it's getting harder and harder to have success in outbound. And there's two ways people react to that. One is driving volume upwards through AI slop and just spraying and praying. The other is actually getting super methodical and targeted about how you approach these outcomes. And we're finding the people who are spraying and praying are falling into regulatory headwinds, spam issues, really not driving a lot of value. And it can be very frustrating. The ones that are seeing most success are really thinking about how do I really hyper-target the right company and the right people with the right value proposition at the right time. And in order to do that, what we found is that you need to have access to a lot of different fields about a company and a person so you can intelligently identify the right people to go after. And that's been the biggest lever across our network in the last, I call it, year and a half for success. And we have case studies with many businesses where they have large SDR teams and the SDRs were manually going into ZoomInfo and filtering or going into Apollo and filtering. That took 35% of their time. They'd maybe be starting to use ChatGPT or Claude to draft some messages. But what we can do is we can, essentially using our models, we can train for your company a unique value prop that understands what's going to work, what's not.
00:18:39 · Daniel Saks, Pablo Srugo
We can prioritize who you should target and then we can have your reps reach out to the high value targets, call it with a phone call or even have your execs go in person but then automate some of the lower tier list. And that's driving 2 to 4x connect rates and meetings booked rates across our network. So seeing incredible value in terms of the time savings so reps don't have to list build anymore or manually think about a lot of the email and LinkedIn and that way they can be more focused on doing what they love and do best which is being in front of customers. And we're seeing the killer outbound teams be people who are very calling centric or in-person centric and they're trying to use GTM engineering via Landbase and Claude or alternatively to us, maybe a Clay to essentially automate a lot of these workflows so they can be very targeted in who they spend time with. [00:19:31] S2: I was going to ask you specifically about Clay like their ripping I mean they just raised like a $7 billion valuation seems to have been what do you think they got right right in the world of like ZoomInfo and Apollo and all these other players like what did Clay get so right in your opinion?
00:19:31 · Pablo Srugo, Daniel Saks
[00:19:31] S2: I was going to ask you specifically about Clay like their ripping I mean they just raised like a $7 billion valuation seems to have been what do you think they got right right in the world of like ZoomInfo and Apollo and all these other players like what did Clay get so right in your opinion? [00:19:43] S0: I think the key thing was there was no infrastructure to be able to aggregate different data sources and therefore orchestrate or automate capability. So in the past you had to have a user going into ZoomInfo or Apollo with a cursor and like make their own judgment on who they go after and therefore it's just a really manual fragmented process and no one data vendor has the best call it access to understand who you should go after and to have the right contact info. So I think what Clay got right was really leveraging the creativity of sellers in a way that they could act as engineers to orchestrate data in new ways and I think Kareem was like super ahead of his time it's amazing what they did because they really started this pre-AI but were able to leverage a lot of the AI to take it to the next level.
00:19:43 · Daniel Saks, Pablo Srugo
[00:19:43] S0: I think the key thing was there was no infrastructure to be able to aggregate different data sources and therefore orchestrate or automate capability. So in the past you had to have a user going into ZoomInfo or Apollo with a cursor and like make their own judgment on who they go after and therefore it's just a really manual fragmented process and no one data vendor has the best call it access to understand who you should go after and to have the right contact info. So I think what Clay got right was really leveraging the creativity of sellers in a way that they could act as engineers to orchestrate data in new ways and I think Kareem was like super ahead of his time it's amazing what they did because they really started this pre-AI but were able to leverage a lot of the AI to take it to the next level. [00:20:34] S2: Walk me through this like you're a go-to-market person and you're going to use ZoomInfo or you're going to use Clay and then we're going to bring in land-based but I just want to understand just from that perspective I mean ZoomInfo is you know pretty straightforward probably like Apollo like you just you said filters like I want businesses there's many employees this geography you know this industry whatever and it builds you a list at least that would be my simple understanding I've played a little bit but I'm not an outbound GTM person right so I don't go super deep with Clay though what would be how would that work what's so different about it?