A Product Market Fit Show | Startup Podcast for Founders
00:08:30 · Pablo Srugo, Mark Hughes
[00:08:30] S1: the way? What what what year are [00:08:31] S0: we talking about? 2017. Yeah. So both kind of worked there for a couple of years. He then had his own his first startup, which was a a YC company in summer twenty nineteen, and I had my first business which was a company called Grad Guide which I took full time and raised 2,000,000 against that business and that was a kind of graduate recruitment platform where we would connect students and grads with a mentor a few years ahead of them on the career ladder to help them find their first role, so it was more like a mentorship platform, particularly focused on sales support roles to help people land those kind of entry level jobs. After that then, like I sold that business in late twenty twenty two and Patrick left his YC company. We'd always stayed in touch and like one thing we like, I realized that I didn't have a technical cofounder. He didn't have like a sales cofounder or someone to focus on distribution, so we always wanted to work together. So to be perfectly honest, the origin story, we have a memo about this that I can share called Mark and Patrick Inc, but it was really like Mark and Patrick wanted to work together and build a business. And then the ideation kinda came after. [00:09:31] S1: So But it was the perfect time. I mean, late twenty twenty two is probably like out of all the times when you could just be like, me and you are gonna do something, we don't know what, that's like the best time for that to happen.
00:09:31 · Pablo Srugo, Mark Hughes
[00:09:31] S1: So But it was the perfect time. I mean, late twenty twenty two is probably like out of all the times when you could just be like, me and you are gonna do something, we don't know what, that's like the best time for that to happen. [00:09:40] S0: Yeah. It was funny. Like, I sold the business in November 22 when, like, ChatGPT just launched, and I had to do like a month handover basically to the new owners. And we were kinda just like working out of a working space in Dublin. We got free access to it at a place called Dogpatch Labs. They gave us a free desk, we were just like tinkering on ideas, taking long walks around Dublin in the cold winter, and eventually then kind of landed on the first version of Solid Road and took it full time in January 23, like I said. [00:10:10] S1: What was the first idea of what you decided you wanted to build together?
00:10:10 · Pablo Srugo, Mark Hughes
[00:10:10] S1: What was the first idea of what you decided you wanted to build together? [00:10:13] S0: There was lots of them. I think the first one was kinda like a gonna be a Gong competitor. Again, meeting bots. But, yeah, there are so many that was quite a commoditized software. So we built the first version of that, and then we send the transcript to ChatGPT, it would summarize the call. And even back then, was hilarious that, like, people were amazed by this because this is before the likes of Gong and Chorus dot ai had even implemented any AI into their product. And so people were like, oh, this is amazing. It's like I can get a summarized version of my my sales goal, and we built that on top of I remember it was another kind of YC company. That was kind of the first version, and then that brought us into the kind of role play training when the kind of speech to text has got slightly better. It's funny.
00:10:53 · Pablo Srugo, Mark Hughes
[00:10:53] S1: The the, you know, the thing you're mentioning about sending it to ChadGPT and that kind of being, like, v one of some product, that's why GPT rappers got such a bad name because that was the original GPT rapper. You just do something, like, incredibly simple. You know? Take it to Chad GPT, bring it back, and, like, this is a product. And then, obviously, a lot of those got killed by either an incumbent like a Gong or Chad GPT. And then the thing became, oh, you don't wanna be a rapper. And still people are saying, like, you don't wanna and the reality is, if you think about every vertical AI app, for the most part, like, they are rappers. They are they are rappers. They're just really good, and they embed the workflow and all this stuff. But it's funny to have that in your origin story just because the the time aligns so well. [00:11:32] S0: Yep. We knew, like, early on when we started using ChatGP for the first time, we're like, this is gonna be massive. Like, there's gonna be so many opportunities on the application layer. So, yeah, just kind of started from there. But it's funny, like, if you can remember, January 2023 was a terrible time to raise money. So, like, we were like, oh, Mark and Patrick, both second time founders, like, somewhat successful. Like, this will this will be easy enough. Let's go raise some venture money. And [00:11:56] S1: It should have been the best time to raise money, but, yeah, we had the macro winter of twenty two, and the AI stuff was just kinda starting off.
00:11:56 · Pablo Srugo, Mark Hughes
[00:11:56] S1: It should have been the best time to raise money, but, yeah, we had the macro winter of twenty two, and the AI stuff was just kinda starting off. [00:12:03] S0: Yeah. So, yeah, didn't have success there, so we were fortunate that we got like, we shook the bookers and got some, like, Irish angels who invested enough to, like, pay ourselves a bit of a salary that I could cover my mortgage and go full time with the business. So even, like, we have, like, a rejection wall of all the VC rejections we have in our office in San Francisco now and, like, a quote that says, chips on shoulders puts chips in pockets. But, yeah, it's been a motivation for us because although, yeah, we've since raised venture rounds and got through Y Combinator, the start was a bit tougher to get off [00:12:34] S1: the ground. Oh, yeah. You've raised, what, over $30,000,000 so far? Yeah. So things have changed. Tell me how you get from what you were first doing, which is kind of this Gong like you know, AI version of Gong, let's say, to the first version of a product, which sounds like it was the the Hyperbound, right, the AI role playing. Was Hyperbound around? Was it like a fast follow, or was it just like you were coming up at same time or different times? [00:12:54] S0: To be honest, I think we might have been actually before Hyperbound because then we saw the hype when they they went somewhat viral on LinkedIn and we're like [00:13:01] S1: We were doing this. Yeah.
00:13:03 · Mark Hughes
[00:13:03] S0: We were doing this. I think I think honestly it was it was a really good moment for us though because I realized that we had to move our lives to San Francisco to build a very big business and you could see like my background now. I've got Koi Terra in the background, but that really became a forcing function. So we started with the Gong competitor, as we said, that was getting commoditized. We knew that they would just bring these AI features on. But then when we had conversations because if you look at startups, obviously, they're a connected graph. So, like, looking backwards, it all makes sense. At the time, maybe not. But then it was like, we really we built, like, these kind of coaching and feedback from the the the recorded calls. And people were like, I really value this, like, coaching feedback. That's actually very interesting. Like, the scorecards in Gong are really crap, and I have no way to actually improve. So then we built the role play tool, and we had the coaching and feedback and be like, oh, this is really strong, but the sales leaders or startup sales in particular didn't really care as much. Went into support, that's what we knew better. We'd, like, scaled support orgs using it. Again, really like the coaching feedback enough to say, like, like, if I actually was to say one thing, like, the flywheel we built from the get go was this the scoring the scorecard and the scoring system, the the evaluations that we built, because then it was like, oh, I love this for training, but I'd love this in our production environment as well because QA is still super manual right now, and we don't have visibility into our conversations. So we use the same scoring system, basically, somewhat evolved all the way throughout. And, like, to answer your question about the HyperMen piece, we had as as Patrick had built all this stuff himself, he wasn't using the likes of Vappy, for example, to build the the role play tool. We went out to San Francisco. There's actually, like, an interesting tweet I can try to service it, but of me demoing Solid Road to people at, like, this YC event because Patrick had done YC before, they're like, oh my god. That's so cool. Super early days, really long latency, etcetera.