A masterclass in founder conviction | Eilon Reshef (Co-founder and CPO at Gong)

Todd: All right, Eilon, welcome to the show. We're so happy to have you here.

Eilon: Thanks for inviting me. Glad to be here. 

Todd: Can you explain what Gong is and who it's built for?

Eilon: So, uh, we're in a leader in a category called revenue intelligence, which is software for revenue teams, helps them carry out all of the critical workflows, coach, better manage pipeline forecast, engage with customer. obviously infused by AI. We've been doing this since 2015. So I've been doing AI for a while, making it more efficient and making their outputs eventually higher for the whole rev org. 

Todd: And so can you rewind the clock a little bit to 2015 and just take us back to the very beginning of Gong and kind of what that origin story was like and where the idea came from.

Eilon: I was in a sabbatical. I'd sold at a time another company I had, and I was looking for something new to do. And then I got to introduce to, Amit, who's now my partner and CEO. And he was running a SaaS company by the name of Sisense, a BI company. they had this issue where they had a really lousy quarter and they're trying to figure out what's wrong. And the CRM all showed like good data. There's deals, pipelines moving, except one thing didn't happen, which is deals didn't close. so what they found out is if they were, they started like recording the sales conversations to understand what's going on.

And if they listen to dozens of them, eventually they got lots of good cues. And at the time he was like, there's got to be a more efficient way to do this. I was in sabbatical. I was learning deep learning at a time, you know, DNN, way before GPT. And I was like, yes, we can, we can figure out how to automate that.

And that was the origin story. 

Todd: And so do you kind of remember where you were the first time you heard the idea and what your reaction was?

Eilon: Yeah, I was in a coffee shop. I was, you know, in honesty, I was dating multiple people first of all, I got time. I was in sabbatical. So, and then I met multiple people with multiple ideas. Amit already had a reputation at the time of one of the most, brilliant or successful entrepreneurs.

So it was like, yeah, I was paying a lot of attention and then I liked the idea, I liked the market. So it definitely caught me interested in, in the first meeting. It was in a coffee shop somewhere in Israel, which is where I'm based. 

Todd: Remind us, what was kind of the state of sales technology at that time back in 2015? What tools were people using or not using for sales calls?

Eilon: Right. It's a great question because one of the, one of the things I noticed was there weren't a lot of sales tools. I had this, like at the time I was looking at, because I was looking for things to do, basically companies to start. So I was looking at the marketing landscape and there was like dozens or hundreds, hundreds, literally hundreds of companies, anywhere from campaign automation, whatnot, and sales really had dozens, which is not a lot.

And my thought process at the time was. Sales, it's kind of like marketing. There's no reason to not have as many tools. so there are probably companies that are less kind of known now, like inside sales. com, which was trying to do the sales engagement back in 2015. Obviously CRM was around all sorts of Chrome extensions that were trying to help you, but there weren't any significant tools outside of CRM, I think sales engagement was just starting. so mostly it was just, Hey, put data in a CRM, look at reports. And this is how you run your sales organization. 

Todd: And was there already kind of a pattern of behavior at that time in 2015? I can't remember, were sales calls happening on video at that time?

Eilon: yes, not as much as obviously COVID. It's kind of changed the landscape naturally. but I would probably say 30, 40 percent of calls were, at the time it wasn't even Zoom, right? It was WebEx and BlueJeans and all sorts of names. sales were happening. Obviously, it was more hybrid than now. mostly for mid range customers. You wouldn't call, you know Boeing, you know, just over, over Webex. Now you probably would, but at the time you wouldn't. So maybe like mid sized customers and our assumption at the time, it's going to grow regardless. We didn't know COVID was coming, so I'll be honest here. but we did assume that over time people are going to be more accustomed, more used and more comfortable with using video versus flying around. I mean, this is so ineffective, just, you know, flying and spending a couple of days on a meeting 

 In 

Todd: 2015, I mean, you guys were sort of early to this AI wave, the idea of having AI assist in a call situation like this, that was a pretty new idea back in 2015. What was, giving you the conviction that this could be a differentiator for you?

Eilon: Well, first of all, you never have conviction, right? As a founder, you gotta have self-conviction because otherwise you're not going to start a company. Most, most people are going to tell you it's going to fail, right? This is one of the things as a founder, you're going to realize, you know, most VCs did not want to invest in Gong in the beginning.

They're like, CRM is good enough. It's a done deal. You don't need any more sales tools. So it's kind of funny. what gave us the conviction is two things. First of all, we knew the problem existed. At least in one company, which is Amit's company. So at least you got one potential customer, right? Here's a TAM of one company. And you gotta assume that if there is a company, you know, it was a SaaS company, I don't know what the sales were, somewhere between 10 and 50 million dollars. You gotta assume there's like like minded companies somewhere. And indeed, you know, later on we found them. So we knew the problem existed.

We knew we can in some, to some degree, solve it. It wasn't about AI at the time. I mean. We didn't even use the term AI because people were scared of AI. but it was a lot around more like sales efficiency. How can you tell what's working? Lots of sales business value proposition that we're pitching. 

Todd: And you mentioned finding like minded companies and you felt like this problem was pretty widespread. Did you do anything to validate that this was going to be a big enough market to support a venture scale company? Or was it sort of just implicit, like every company has salespeople, this market will be big enough.

Eilon: I think it's, it's sort of top-down Yes. You sort of have how many salespeople are going to be are in the world, somewhere between five and 10 You multiply this by, you know, pick a number, a thousand bucks a Salesforce charges more or less. You have, It's a VC size market. I think what was more interesting for us at the time is can we find a beachhead? So it was less about, Hey, can we eventually grow and become a, you know, have a fortune 50 customers? Of course we eventually were successful. We will get there obviously. Now we do. but at the time it was more like, can we find a beachhead? And then it kind of led us towards like, Hey, let's find those companies that. You know, are kind of more lend themselves to be early adopters versus the ADPs of the world or, you know, the, the Amazons of the world or the Googles of the world.

Todd: So when you were just starting out, who did you think your beachhead was going to be? And did that end up becoming the right beachhead?

Eilon: Yeah so when we validated the, market it was easy to validate we said, let's kind of focus on technology companies. again, we kind of knew the market, we knew how to approach them, how to sell to them. So, and we also had, of course, you know, personal relationships. So we kind of assumed this could be the beachhead and indeed, you know, probably first, I'd probably say 30, 40 customers, probably most of them were in this space. initially we were like super focused. I think we're probably one of the most crazily focused companies in, in early stage. We're basically aiming. software companies located in North America selling B2B somewhere between a ticket size of, 1, 000 and 50, maybe 100, 000 in English over a handful of video systems. You know, there's probably maybe now 10, 000 companies like that in the world, probably five, probably less. Um, but it was the initial beachhead. And then we're going to later we do on many other things, but we wanted to sort of start with, noticeable impact in a smaller category. 

Todd: Let's talk about when you actually started. focusing on building the initial product and kind of that zero to one phase. once you and Amit start to kind of go all in on this idea, how did the two of you delineate responsibilities?

And, and maybe who else was on the team at that time?

Eilon: when we were raising money, it was just the two of us. And initially, you know, Amit's background, Amit was actually an engineer many, many years ago. And then he did marketing and sales. So, he basically asked me, you know, Eilon, what do you prefer product or engineering?

And I'm like, what do you prefer sales or marketing? And then we ended up like, just like he did sales and marketing. I did product and engineering. And later on, I kind of focused on product. He later on focused on being a CEO. So kind of all diverged. so that was the original, original kind of split.

Todd: Talk a little bit about the, very first kind of demo or MVP and, and sort of like what you thought needed to go in it and, and who you got feedback on it from.

Eilon: Yeah. So I think there's two baby types of, you know, what do you call a demo? There's a slide. There is a deck. I mean, you got to have some sort of a deck where you talk to a customer. You want to show them something, even if it doesn't exist. So that was like pretty basic. You know, Amit and I put it together.

Probably Amit did 80 percent before he even met me. I mean, it was, you can. Probably still use this deck today if you wanted to sell a very, very basic version of Gong.

Todd: And so that was before any software was written? You were, you were sort of going out there with a, with a deck?

Eilon: Yeah. I mean, before software, and then Amit kind of forced hand me into, into, I was running kind of mid sized engineering product organization.

And it was like, Hey dude, you're like the engineering guy. You got to write some, something for, the company. So I was like, okay, I'll download an IDE. I'll open an Amazon account, a cloud account using my gmail account. So I put together myself like the first prototype. I kind of forced myself to remember how to code.

And then we put together a very basic software system. And what it did, it basically joined Webex calls as a bot. you'd see like a bot on the call, it would basically do screen recording. And then he picked up the audio and the video and then introduced a third party transcription engine at a time.

So I was, couldn't remember myself, so I couldn't write a transcription engine. And then we would like index it or whatever. And then let you kind of review it and maybe comment or whatever I could do in a few months of like personal work in addition to founding a company. I had one engineer who helped me like as his side gig.

He was, he later became our employee number one. and we took that one to market. So we basically took it to another, another company. and then we let them kind of play around with the software because I couldn't scale it. It was just like recording one call at a time. So the most you can get is six calls a day or five, however many.

but it was still, they found, value in it. So we basically said, it kind of makes sense. And then we kind of started working with others as well.

Todd: You took it to customers that had already seen the deck and were kind of warmed up and then you showed them the demo some months later?

Eilon: Yeah, we pitched to a few companies. Some of them said, Hey, show it to us when you have it. Because it's like, yes, we like the idea. Some of them were like trying to be nice. So they said that and never came back. And a handful of companies said, Yeah, give it to us when you have something ready.

So whenever I had the first version stitched together, I was like, Hey, James, remember you told us that you kind of, want to try it out. Here is, here it is. And, you know, I think luckily for us also at the time people weren't, as aware of, security as there are now. So it was like, yeah, sure.

Bring it on. I'll try it out, which today I think you'd probably have to pass through 50, uh, security reviews. So. Company was Act-On Software, so Act-On Software, they did email marketing. They found it very useful, so it gave us lots of positive energy.

Todd: Do you remember what the most positive reactions to that first demo were and what the most kind of negative reactions were?

Eilon: The negative reactions are, have always been, by the way, until today, because we still, part of what we do is, is ingest and analyze conversations. So the negative thing has always been, hey, our sales people are going to want to be recorded. Like, oh, big brother, is anybody going to listen to my calls, et cetera.

And like, dude, the managers don't have time to listen to your calls unless you ask them to. So it's not like, nobody's monitoring you, but it's always been a concern. And I think, so beyond this concern is the fact that we're able to provide value to sellers. Okay, I'm willing to pay this price of being quote unquote monitored because I get so much value out of it. So that was always the negative. The positive was just the dream of AI telling you before it was called AI telling you what works and what doesn't. It's even in very initial versions, the system would have told you, Hey, you're talking too much. And like half of the people just talk too much.

I can't believe it that I just spoke 70 percent of the call. I should probably be speaking 45%. And that's the first thing the AI told us, Hey. Just look at your talk ratio and there's no excuses, right? If your manager told you you're speaking 70%, you're gonna be like, no, you didn't time it right.

It was necessary. The customer just didn't respond, whatever the excuses are. But when software shows you 70 percent and then 65 and another call 80, obviously you can, you improve yourself very quickly.

Todd: Did you find that you had more, interest and pull initially from the sellers or from the managers? 

Eilon: It was both. I can tell you that I remember this one onboarding class a couple of years later, maybe three years later, when we had two people coming from a customer LinkedIn. One of them was an AE and one of them was a manager and like the AE told me, Oh, now that I'm in Gong, it's the first time I realized the managers get value out of Gong.

I was like, Oh, that's interesting. And then the manager told me it's like interesting because the first time now that I'm in Gong, it's the first time I'm realizing AEs get value out of Gong. And it was the same, the same class. And I think just everybody assumed it was for them. And we're like, okay. Fine.

Sure.

Todd: So how long did it take you to build what you would consider like the first, you know, solid beta version of the product?

Eilon: I was lucky in that I was able to bring a whole team, what now is being called maybe a pod, for my previous company, because we had sold it a year earlier and the guys were bored. The people were like bored and being part of a bigger company. So we raised money in October of 2015, all of them joined like mid November.

And we had probably something I might call early beta somewhere in January. So it was like two months into it. And then we kind of threw a handful of customers on it in January. So that was way quicker than maybe I would have expected, but it was, it was nice.

Todd: And how did you think about what goes in the beta? What's in versus what's out given that you launched it pretty quickly?

Eilon: We had this very strong hypothesis that just giving people visibility into what's happening within a sales organization is super important. So we had debates. I even remember one of our investors telling me, you got to have your own speech to text engine. I'm like, sure, of course, but I'm not going to put it into a beta.

And he's like, well, that's not affordable. You're going to pay too much. Even now, you look at like what Google charges, it's an arm and a leg. Right. but there's no way I'm going to build it. He's like, how are you going to do unit economics? I'm like, yeah, I'll tell you about unit economics when we get to round B or C right now I want to improve product market fit.

Right. So the idea was like, keep it very lean. Almost no sophisticated stuff, but make sure you get those precious sales conversation in front of the right people. So we kind of had recording, we had transcription, we had commenting capabilities so that you can bring on more people on board and a little bit of this, like very lightweight AI around, you know, talk ratio kind of thing.

And the hypothesis was. Just giving it, putting it in your face it's the first time you're seeing yourself as a seller, just how you perform. And then for a manager, it's the only way for you to ever kind of see your team without being on the calls. And that hypothesis kind of proved itself. So, we didn't need to iterate much in order to get quote unquote product market fit.

Todd: So it makes sense to me that you were, trying to get this thing together quickly and trying to really test for customer love prior to thinking about things like unit economics, efficiency, stuff like that, were you, so you weren't worried necessarily about what it costs to produce, but were you worried about what you could and charge for it? 

Eilon: I think initially we weren't even worried about what we're going to charge for it because we had a very, this grand plan of what Gong does now, which is, hey, this semi autonomous revenue system is going to help you do forecasts. It's going to help you manage pipelines. So we knew eventually value is going to accumulate.

so our biggest concern was, is there value to the people or first of all, is there value to sellers because we are asking them to record calls? I mean, it was automated, but, you know, you agreed to give up some of your quote unquote privacy. And can you give value to the organization? So the assumption was we're going to price somewhere like, you know, at the time the benchmark was CRM Salesforce.

So it was your stage, you have a, seed company, you're not going to charge as much Salesforce, maybe half of the time. so we knew there was going to be some value. I mean, if we can get there, but what we want to prove is there is value, there's user love, there's buyer apprehension of the value, these kind of things.

Todd: what were the first signs of that to you? did Gong pretty much work right away? you felt like you had the signs of product market fit right away or was there iteration required? 

Eilon: In hindsight, yes, of course, we're in the trenches. You're not, you know, I wish I was. Um, maybe kind of more self aware or more maybe optimistic, but you're always like, Amit and I both kind of look for negative signs versus positive signs or like, Oh, this is not working. This company is not implementing, blah, blah, blah.

So always looking what you can improve, which is not a bad idea, but sometimes you're like, you kind of lose a little bit of this celebration mode but yeah, the first customers used it. In hindsight, probably the first indication there was a strong product market fit were probably 9 out of 10 complaints were how this, how come you didn't even record this call?

This call was not recorded,

which ironically, Amit, when he had this vision for Gong, he was like, let the first value of this is for sort of the company to know what's going on. So even if you record 50 percent of the calls or 60, you know, you can do the analytics. It's going to work the same way.

It turns out sellers and managers were like, you know, now I'm relying on this to be my ears and eyes and memory and notepad for all of the calls. So in hindsight, again, getting so many complaints about our system, not recording all calls, which it wasn't even designed to do in the sense of we weren't worried about bugs or timing or whatnot, should have given us better indication that this it's like becoming a critical tool versus just a nice to have.

Todd: yeah, that's an amazing sign when, customers become reliant on something, you don't even really realize it. And then a call fails to record and they're like, Hey, what's going on here? Like I need this. 

Eilon: Exactly. Um, so I think we all appreciate, you know, sort of like criticism is better than if people ignore you, they probably don't need it. If people start complaining, that's usually a good sign.

Todd: Eilon, do you have a, I'm curious how you think about design partners and this is maybe for the founders in the audience. What makes a good design partner? How do you choose the right design partners and how do you work with them? 

Eilon: It does depend a little bit on the stage of the company in my mind. I want to maybe start by saying at Gong were like maniacs about using design partners. So each one of my PMs, we got maybe 20 25, PMs now, more or less, at the size we are thousand people organization ish. each one of them is working at any point in time with probably somewhere between half a dozen to a dozen design partners.

So if you think about just like how many design partners does Gong have at this point in time, 2024, that's quite a bit. And we haven't like changed that. So. We're strong believers in design partners. usually you want to have early stage. You want to have the innovator types. The people who are like, of course I can do this.

It's like, oh, this is going to change my life. This is somebody who's like maybe a little bit overexcited because they see the potential, not what the tool does today, but what it could do for the organization where we are right now as Gong We also look for other types of customers. The more programmatic ones were like, you know, can I can I configure it?

Can I get the most out of it? Does it actually work versus does it promise to work? But in both cases, it should be individuals, who basically believe in the value, they're willing to invest some time. and they kind of align in, thinking how this could change your business and willing to give you feedback. They don't just like put it in and expect it to work kind of thing because I don't get any value out of it. 

Todd: This is for B2B founders, but do you find that founders should make a conscious trade off of going for these design partners who are really enthusiastic? Forward thinking, even if they can't pay you very much.

Eilon: I think early stage, I would not worry about payment as a key thing. I mean, there's always a question of, you know, can we, can you charge for it? And there's many ways to sort of find like, maybe proxies for value in some way or form. But I think if you sort of insist on paying, you're basically kind of putting, asking people to not just like put trust in, this is a good work, but actually quantify it.

And they haven't seen the value yet. So maybe for some systems, maybe if you can do like direct say, cost saving on ads, right? This is something people are going to be willing to pay for because you just save them money. But if you're asking them to invest in something new, they've not, they don't even know how to associate value with it.

I think as a founder, you first of all, want to see, you know, A is their value. B is their impact and C Probably most importantly, if there is impact, are those people willing, are the kind of people who are going to be eventually willing to speak up and say, yes, I've installed this software. It made this this impact for me.

And that gets you the next wave, because if you don't get that, it's very hard to kind of get to the next wave.

Todd: So we're all, you know, your first 10 or so design partners, when you were getting started, were they all free to start and then eventually you got them to pay?

Eilon: Yeah, yeah, yeah. So we had like 12 design partners. We got them through network and a consultant we had worked with and then we all, it was all free and some stage. They stopped complaining about calls not recorded because we kind of fixed most of the issues. And then it was a little bit of quiet.

So we're like, okay, if people are just using it and not complaining, we maybe ought to charge them something. So we called all of them and said, Hey, beta is over. We're starting to charge something. That was six months into the company. And then 11 out of the 12 paid, which to your previous question, probably should have been a good, a good timing to say, Hey, we have a good product market fit, but we're like pretty casual about it's like, Oh, that's good.

Now, how do we scale that thing versus celebrating the moment? 

Todd: Okay. Those are two real milestones though. The first milestone being when people started to complain that the calls weren't recording and then the second being converting 11 out of 12 design partners to paid. That's solid.

Eilon: The only company out of the 12 who didn't buy this year was leaving. There's still a customer. So it was like they bought a year later. 

Todd: Basically 12 for 12.

Eilon: Well, some of them are not around. And then the funniest story is one of those 12 was basically, they actually didn't want to buy it. So they're like, we shut them down.

A day later, Amit gets a call from their CEO. It's like, why the F did you shut us down? It's like, Oh, the CRO told me that you don't need it. He's like, What is he talking about? We use it all the time for blah, blah, blah, blah, blah. Could you please turn it off? So up until now, we have this feature. We can kind of deactivate and reactivate the customer and keep the same state, so to speak, because of that customer.

But, you know, one of the 11 that actually bought, but a day, a day too late maybe.

Todd: And so just shifting topics a little bit, how were you thinking about fundraising through this journey? It sounds like you raised at the very beginning based on the idea and then you're starting to see momentum and you're starting to see some of these design partners convert. did you then fundraise again at that point?

Eilon: A little bit later, I think we had, I think we raised the time 6 million, which wasn't huge, but at the time was decent. So we gave us more leeway to kind of get to the first, I don't know, it was like a million dollars, maybe. so funding came a little bit later, but even raising an A round was not easy.

people were still skeptical, you know, the same concerns are, people aren't going to want to be recorded. Is the CRM enough for everybody? So A round was still tough. I think beyond that people started seeing a lot of momentum. But until we got to this lot, lot of momentum, it wasn't easy. And people weren't haggling or like, Oh, we're going to fund this at like valuation X versus Y, which obviously a few years later, it kind sounds a little bit funny.

Todd: I'm a little bit surprised about that, that the A was challenging because you had pretty much all of your design partners converted to paid. You had people telling you they can't live without this product. They have a CRM and they're clearly also buying Gong. Do you remember more of that feedback?

Eilon: I think some concern was, you know, is this a separate category, which is a valid, always a valid concern. They start something where is it going to be. Some people were confused about this being having overlap with system for call centers, which were also analyzing calls, maybe they entered this market.

why would somebody selling to call center ever get to B2B kind of companies, but it wasn't obvious that they wouldn't. I think some of them was like, we're concerned about the size of the market. You asked before, you know, what percentage of calls are being records like, Oh, it's only for inside sales.

Okay, well, I think we were pretty confident that over time, you know, a huge portion of conversations are going to be digital in some way or form. But again, that was our conviction. Not all VCs are going to share the same view.

Todd: Let's talk about now kind of the next phase. So you've sort of established, you know, these signs of early product market fit. How did you go from your first dozen or so customers to your first hundred?

Eilon: So that's when you start building like a real sales team. So we hired this one SDR. Amit and I were still based in Israel. So we hired this like one SDR in Israel who basically was SDR for Amit. I mean, it was like the sales guy, right? And then subsequently we started hiring real people, like real sales people out of Palo Alto, and obviously kind of more closer to the market. A small team, you know, COO, and then one first account executive and then first customer success person, and then maybe first support person.

So we kind of obviously kind of gradually one by one kind of starting assembling the team, the more obviously customers we signed, the more confident we felt about kind of growing it a little bit more.

Todd: And so what does that transition look like for you guys as founders? Because at the beginning all sales are founder led, right? And then you start hiring out this team. I'm imagining a pretty experienced team. how do you let go of the founder led sales kind of over time?

Eilon: Well, first of all, I remember Amit did a sales and marketing and I did a product engineering, so it's very easy for me to kind of let go. I probably closed one or two deals, not more than that in all of the, relationship kind of stuff. But I think he was eager to sort of have some of this being scaled.

So I don't think anybody's attached to this. you know, just closing every single deal. It's something you do because you got to do it, but obviously you're happy if somebody else does. So we were very happy. I was like excited when the first, he sells his first deal. That's a big thing for a company, right?

it's not like me selling. It's somebody who's been trained and can convey the value. There's some repeatability element to it.

Todd: when you got, past 20, 30, 40, 50 customers on the, on your way to a hundred, was the basic product still the same or did the product kind of expand in a lot of directions?

Eilon: I mean, Gong has been traditionally kind of very fast and sort of expanding our, our solutions. So yes, initially it was, Hey, it's going to record your calls. And then we added the whole layer of coaching because that's the first thing people were looking for. It's like, can I leave comments to my reps?

Can I see aggregate stats of like what all my people are doing? Can I see trends? Can I search across all of my calls? Over the first year, it already became this Oh, can I be the library for calls for onboarding? So probably the first couple of years were focused a lot around, It was also aligned with like our customer base was growing and then they were growing as companies.

Many of them were tech companies. So how can you help me make sure my reps are performing? How can you help me scale my team? How can you help me onboarding new people? Nowadays, it's like a secondary thing because people don't onboard as aggressively as they might have done in 2017 18. It's still a very popular use case with Gong. So those layers were built over the subsequent couple of years.

Todd: And how much of the roadmap was you and Amit's kind of intuition and feelings on, what should be built versus your customers and what they were saying. It sounds like so many of these ideas were just sort of like coming out of your customers and you guys were racing to keep up with what they were asking of you.

Eilon: I think the strategic roadmap is maybe 80 percent quote unquote vision. The tactical roadmap is probably 80 percent customer led meaning. Customers were asking us about something that we nowadays call scorecards, which is ability to kind of create rubrics where a manager can score individuals. And we had thought that that's a call center thing, but turns out even in B2B, you want managers to be accountable to a given methodology, and you want them to sort of go by a certain structure of how do you evaluate somebody's call to help them, not to score them per se.

We didn't know that was a thing in the B2B business so we built it because customers were like, Hey, this is how we'd to run our coaching program. At the same time, this whole notion of where Gong is maybe now, which is the system that semi-autonomously understands what's going on, recommends next steps, writes emails for you, which now is possible with GPT.

It wasn't possible at the time even. helps you forecast, helps you understand which deals to focus on. That was Gong led. So when we built this functionality, Many, many times it was a head of customers asking, and then when the head is coming, they're like, Oh, shit, this is cool. And then a flood of requests came in.

It's like, Oh, if you have this, could you add this, this and that? So I think still up until now, I think the next thing that the big thing that we're developing in many cases is market category led, how does the category going to look in a couple of years? And then within this category creation process, it's what are the pieces is what customers tell us that they want more.

Todd: I think it's interesting this framework of you've got, there's a strategic portion of the roadmap in which 80 percent of it is driven by founder vision, a company vision. And then there's the sort of, I think you call it the tactical part of the roadmap, which is 80 percent kind of customer driven. What is that, how does that actually map if I've got a hundred engineers, how many of them are working sort of on the strategic stuff versus the tactical stuff?

Eilon: We have various frameworks. I probably want to say at any point in time, it's first, it varies because sometimes you overinvest in a new product. But I'd probably say in most cases, you probably we probably do like 50 percent in a core product, 25 percent in new and then 25 percent in net new, which is call it strategy led. Some years maybe we've done a little bit more of net new and some years we're like, okay, we got to catch up to stuff we've already built. but generally it's maybe 50, 25, 25 ish.

Todd: so it seems like in the first, you know, two, three years, you had really nice growth. Do you remember there being sort of like the next big, growth moment for Gong and a milestone that where you were like, Whoa, this could be really big.

Eilon: Yeah, probably if I remember correctly, I'm never sure about the numbers cause it was a while. I think the second year, like the real first full year of selling, selling business were like at 2 million ARR which is not bad for your first year. And then second year was like $9m And that was a big thing, like moving to from two to nine, that was okay, sh*t this is exploding.

also meant, of course, who do you hire? The third year was also pretty, I think it was like three X. but this is true hyper growth. It's not like the word, but it's you feel it inside because, you know, for every customer you have, you're going to get 40 to the four more.

So it's pretty crazy.

Todd: And when did you start thinking about kind of going multi product and, and talk a little bit, I think it's interesting Eilon that you, you still have design partners, it's still a part of the Gong way of building. Talk a little bit about going multi product and then finding product market fit for the new products.

Eilon: Yeah, I would probably lose an arm before I lose this. The concept of design partners. So it's definitely, uh, I think how people should be building every one of our new products is like built with people in, maybe they think of it as beta, but it's essentially design partner program. With the new products that we've developed, there's probably, I mean, depending on what you call product, cause it's like features you might label this product and there's SKUs, that you sell separately. The two SKUs we've built after we built our core Conversation Intelligence product as on our route to what we now call Revenue Intelligence were, a SKU on forecasting process and a SKU on sales engagement, which is interacting with customers.

In this case, it wasn't a brand new category. It was existing category. we've incepted a product that we believe is you know, better than incumbents and in certain ways has many advantages. Obviously for V1 is never going to be, feature parity to people who have been around for five, six years, but has always those kind of AI nuance or understanding recommendation stuff.

That is not just data entry. in this case, I don't think you're looking for product market fit because in the end of the day, product market fit is for a category, right? Do you need forecasting software? Well, I do. I already have one. So it's more about. call it like maybe product completeness questions is the product complete, does it solve the business problem first of all? And then you know the second question is, can it compete successfully with maybe other players in the market? So building to call it like outcome to make sure it provides the outcome is anyway easier you work with the design partner you build it together you see that you get the outcome.

Billing to make sure you can actually compete successfully win 50, 60, 70, 80 percent of the deals is much harder because over time you find out what is customer number 37 want and need, which you might not have seen with the first batch of design partners.

Todd: And so what are you looking for then as indicators of success? Is it metrics based? Is it, you know, sort of anecdotal talking to customers?

Eilon: So if it's a new category, the first thing we kind of launched, it wasn't a separate product, it wasn't a SKU, but it was the whole idea of pipeline management. we basically asked them this question, that was kind of popularized by Superhuman, which is, how disappointed you're going to be if we took this away.

So this is more like a product market fit kind of question, which indicates, you know, is there an early batch of customers who is very disappointed and what, big is that batch? Right. I think for the new products in existing category, and let's take Gong Engage, for example. 

It's a product in the sales engagement category, helping, um, sellers interact with their customers. There's other players in the space. What we're looking for is of course, just business considerations such as, you know, what is our win rate? What is our average revenue per user, right? And then eventually what is the customer satisfaction?

Call it NPS or CSAT or whatever metrics, you're seeing. We try to also look at outcomes, right? Have them been able to achieve goal X, but I think on the way to achieving goal X is like, how are you happy with, you know, is it set up correctly? Is it helping you do your job? Is the user interface friendly?

Does it not crash? All sorts of things to, on the road to actually delivering the outcomes.

Todd: Do you remember when you were asking that, how disappointed would you be if this went away question, kind of what the responses were like?

Eilon: The responses were very high, meaning I think the number was very disappointed. It was like, I don't know, maybe. I think Superhuman's at 40 percent is a threshold of some sort. I think ours were like 80 maybe. it was nice, but I think it was, I think we over read into this because I think you asked the question of people who actually use this piece of the product.

So you very early see that there is a product market fit to a certain portion of the population. What it doesn't help you do, and it took us some time to realize that it doesn't help you understand what is the reach. So is it like 10 percent of your I mean, you see it separately, but not this metrics is it 10 percent of your customers even use the product enough to tell you to answer or is it 20 or 30 or is it 80?

Because once you go out of a single product or even a single use case, it's always a question of, you know, can you impact the whole population or you developing a feature or product that only applies to certain portions of your customers? And then it's, the, uh, process is harder or just the uptake is not as big.

Todd: Yeah, that makes me wonder, when you're launching a brand new product, do you think of it as like, what percentage of Gong's current customer base will adopt this new product, and that's what we're optimizing for? Or is it like, Hey, this new product will bring us into other segments and we're actually looking for net new customers with this product.

Eilon: Really depends on the product. So the first, product forecast was definitely intended as a cross sell motion. So it was can we get to X percent of Gong customers using it within the first year, and then for second year. Gong engages also, initial motion I think right now is as two thirds or 75 percent of customers are existing customers.

So it's still a lot of cross sell, versus new, business. But it might be a land product one day. In our queue, we have other products which we believe are going to, get us in into customers who might not have enough, or might not be getting enough value from Gong's existing products. To give you an example, right, let's assume somebody does not, cannot legally record any calls. So the value, they would still Be getting value out of forecasting or sales engagement, but the core conversation intelligence is limited. We do have customers using us that way, but it's limited. There's a bunch of things we're doing to say, hey, if we ever go to more regulated industries or, I'm gonna, maybe as a pun, right, Germany,

they're never gonna record calls because they're Germans. so maybe there's a product that's gonna work well for them as well. 

Todd: I want to shift a little bit into AI because again, I think, you know, in 2015, I think you guys were really early, to this market, which is obviously the thing that everybody's talking about now. Do you remember kind of when you got started, was there anything unconventional about how you were building and using AI in the early days?

Eilon: I think first of all, unconventional was just about bringing AI to business users. Most people when they were talking about AI was this could be a predictive model for placing ads or this could be something that runs in the background. People weren't necessarily even expecting, AI to tell them something as business users.

So I think that was the first thing. And there was a lot of debate and thought process around how do you bring it? Because AI is going to make mistakes. Every AI is going to make mistakes. Humans make mistakes, right? So how do you even like present stuff to people that is known to not be 100 percent accurate?

So there's a question of, you know, what is the what you I for it to say at the system might say, Hey, we think this is what you should be doing, but hey, just verify before you do. So there's a lot of, non standard way to approach customers. even you tell somebody, Hey, this is a call. You should get a follow up on this call, right?

Should you follow up on every call? there's the tone of the voice and tone and like we wrote this email for you. We probably not the email you want to write, but you can probably get it fixed within a couple of typos. It's like, Hey, this is AI generated. It's like, oh, and please review it before you send it out.

So I think the mechanism of communication to customers is still being refined, but that's been a lot of investment on our end..

Todd: Do you notice that there's a big difference in how you're talking to customers now about AI and how you're integrating it into the product and sort of just the education of the market that's happened over the last eight years? Like, do you talk about it differently now than you did back then?

Eilon: Yeah. At the time we just didn't say AI, we would say, hey, this is like revenue intelligence or conversation intelligence. And then AI people even know, I think they thought it was an action item. Nowadays you start a conversation like, hey, you want to introduce AI to your revenue organization.

This is how AI affects your revenue organization. You know, admittedly there's much more AI in Gong than it was, or used to be like six, seven years ago, but people are much more receptive. They know they got to use AI. So it's more like, what do I use? Obviously people think that, you know, GPT is AI, but GPT is not going to help you sell more.

So, so they're kind of thinking, you know, what is in this promise that I can bring to my revenue organization, it can scale and can deliver outcome, connect to my systems and so on.

Todd: Are there developments over the past two to three years, you know, related to large language models that you think are really interesting, relevant to Gong, things that you've been keeping an eye on?

Eilon: Think the whole LLM piece is very relevant to Gong. We've taken it in directions that are not necessarily like the GPT kind of conversational interface, which I think is nice. But it's just one angle. I'll give you an example of how we've kind of put it together. So, one of the things revenue leaders want to do is also quantify behaviors or not just take summarize a call for me.

It's like, are people following my message or are our customers asking, you know, about whatever it is integrations. For this to happen, you actually have to train an AI model to correctly identify the thing you care about. So it's not just like, ask GPT, you know, are they following the message at GPT is going to say what message, right?

There's no idea what your message is. Right. And then we build this I call it like Tinder, like interface where it basically kind of gives you option. You're like, yes, no, yes, no. And within a few examples, it kind of learns that stuff. So I think where my mind is at is less kind of the core technology or the using RLHF or the using, you know, it's going to open AI or versus Mistral.

Like foundational models are kind of awesome. I follow them all the time. We kind of pay a lot of attention to them, but I think what we're more concerned with is what is the, how do you bring it into the world of the revenue organization in a way that brings image value doesn't have to be generative.

The example I mentioned is 50 percent generative. First of all, it identifies, then it kind of writes down the explanation. some others are generative, you know, take right in last week's update for me, right? It is generative, but again, it's less about GPT and more like what data do I use? What tone do I use?

What needs to be in a, in a, in last week's summary versus, you know, GPT, could you please take this like five calls and try to write something, right?

Todd: What do you think, B2B sales looks like, you know, three to four years from now.

Eilon: Forrester has this nice stat where I think sellers are spending 23 percent of their time selling. 23 percent meaning 77 percent is not selling and they kind of include in the 23 percent not just FaceTime, just writing emails and other things that are productive. So 77 percent is almost non productive.

I think that one is going to go down significantly. We internally, we call it drudgery. So things like updating CRM fields so that the company can eventually analyze them. AI should be able to do this. share information, write summaries, write account summaries, be in debrief meetings, account handoffs, account handover, account strategy documents, all of that AI should be able to do.

And the stuff people hate to do, right? Are you writing emails, by the way, you're still gonna edit them, but not write them from scratch. So a lot of the drudgery stuff that people have to do, it's going to leave salespeople with the good stuff, the fun stuff, which is building relationships, which is what they like to do.

So I don't think they're going to go away. I mean, people are like, oh, can a virtual rep call you and sell you something? It's like, I'm not going to pick up the phone for what, in the next 10 years, but certainly make their work much more efficient.

Todd: So Eilon, zooming out a little bit, if you were to give founders advice on finding product market fit and the advice that you actually used yourself, what would some of that advice be? 

Eilon: You know, I would always urge anybody to work very closely with a customer. Get in bed with them. You have a software idea, you go in and implement it with the customer to make sure they get there, like fully in love with it, we call it internally raving fans, make sure the customer is like super happy about it.

And then scale sometimes, you know, maybe also what I've done in my previous company is kind of a little bit, you know, got to an okay version of a product and then kind of started sending it to more people. I think just going in bed with a handful, a small number of customers, making sure they're super successful and super enthusiastic is a much better way to kind of know that you've got something right versus something that's okay ish.

Todd: You mentioned your, previous companies. How did your experience starting Gong differ from your first startup Webcollage? Like what parts of company building you know, did you expect to say the same and what surprised you? 

Eilon: First of all, I do recommend to all founders to start with a second gig. It's much easier. I wish I'd followed this myself.

Todd: to start with the second one? Of course, it's always easier, right? I probably made so many, when people ask me about mistakes in Gong, we did make mistakes, but I'm like, not so many because I made so many of them.

Eilon: I'm like, my bag is full of mistakes in the first one. So you kind of learn from mistakes as well. Hopefully not just from mistakes, but I think in the first company, we try to scale too fast. So, we had this one customer and we're like, okay, good. Now we're going to hire six salespeople and sell.

Which is ridiculous because you want to make sure that you're entrenched with the first few customers, as I just mentioned. So I wish we had set a time like more founder led for the first 5, 10 customers. Probably would have been in a much better shape because sellers are. Not as, successful or not as knowledgeable and then you don't get enough feedback.

So scaling too fast is probably, uh, tempting. Sometimes VCs push you to do this, right? But, you know, first time they push you, second time it's harder for them to push because you've been pushed so much before, so.

Todd: A couple of wrap up questions what is your non obvious founder's superpower and how did it help you build Gong?

Eilon: I'm probably more on probably top 10 percent of people in terms of curiosity. I truly care about lots of things. I like, so obviously gives me a very easy time in learning new stuff. when I learned deep neural networks, it wasn't because I need this work, it was like, Hey, this is back in 2015. I was like, Oh, that's cool. I got to learn this. And, you know, I kind of took a few data science courses, you know, being, being a product, I don't have to do this. It's like, Oh, this is cool. This is interesting. so it does help. I think it helps a lot because it kind of makes it very easy for you to learn.

and that's true for industries. It's true for customers. It's true for technologies and it really helps you as a founder. I remember going to this fortune 100 customer that was my previous company. Sales call and basically customers come in and like, okay, so can you tell me a little bit of your business and you know, they tell whatever they do and I'm like, oh, can you tell me this. For a whole hour, I basically interviewed him about all the stuff that they do and then the hour ends and the guy kind of escorts me to the way out, which at the time everything was physical and it's like, wow, that's the one of the best calls, sales calls I've ever had.

I really didn't tell him anything about a product. He was like, you understand our business so well. Of course I did. Cause I kind of listened to him, but I was, it wasn't because I wanted to do a good sales call. It was just because I was curious about their business. It was a retailer I wasn't familiar with, and it really helped me kind of understand their business.

But it was even good as a sales mechanism, but I think it comes from true curiosity because you can fake it. You can fake it all, 50%, but if you're not truly interested, it's going to much, much harder.

Todd: Any favorite books or resources that you'd recommend for other founders?

Eilon: So there's, I mean, there's numerous books across, I would probably every founder should probably be reading the, the more like the classical kind of Crossing the Chasm by Geoffrey Moore. I mean, if you're a founder and you have not, you should probably read it. If you've read it, maybe again, is what we've done with Gong.

I mentioned before, we kind of start with a very small set of customers and then sort of the bowling pin kind of started moving to other directions. So recording phone calls, doing emails and other things. First company did it completely wrong. I read Geoffrey Moore's book. We had this one customer in the cosmetics industry.

I was like, ah, that industry is too small. Let's go to different industries. And maybe two years later, we're like, hey, we got those 10 cosmetics companies and there was a great beachhead to other manufacturers. And, you know, I figured I was going to be smarter, but, that's probably where I would start so. But you know, being in this forum, probably worth mentioning First Round is one of the, uh, blog posts I am actually subscribed to, not just read, subscribe to and read. So kudos to you for putting it in place.

Todd: So it's 2024, you know, about nine years after the founding of Gong, uh, what's next for you and the company? 

Eilon: Well, for the company, it's always how do you continue to grow, make customers successful. For now, what's top of mind for us is making sure we can be the one stop platform for all of the revenue organizations. So, if you look at our typical customers, we serve salespeople, sometimes in the tech industry, customer success managers.

And over the next couple of years, plan to kind of provide a system that's going to help say everybody, account managers, post sales professionals, it's going to be this one system that serves them all using unified analytics platform, unified workflows, and many other things that I think they don't get today from any vendor or any system.

Not a CRM system, not a Tableau system. So this kind of unified system for everybody who's touching the customer.

Todd: Uh, Eilon, thanks so much for being with us here today. You were a great guest. We really appreciate it. 

Eilon: Thanks for inviting me.