Bringing a Generative AI Product to Market: RJ Talyor, Founder CEO, Backstroke and Pattern89 (Part 3)

Sramana Mitra: Two questions. One is, who came up with this technology?

RJ Talyor: It’s a big combination of things. My co-founder, Jeff Cunning, and I kind of dreamed it up.

Sramana Mitra: That’s the management part of it. You positioned the product and you kind of envisioned the product. But somebody still has to do the algorithm.

RJ Talyor: Yes, exactly. So, we worked with our engineering team. We had a great engineering leader named Matt Brown, who was early on in the business. We worked with James Winans who ended up being our VP of engineering, and then our data scientist Charlene Tay.

James, Charlene, Jeff, and I kind of hashed it out over time with customers telling us what they loved or hated about the product. For the first year and a half, we were manually running statistical spreadsheets to understand what worked. As we found signals within the performance, we codified it and made it the product.

But literally every night, it was Jeff and I rerunning the numbers, recalculating, and then checking with Charlene, “Does this look appropriate? Does this pass muster? Is this statistically significant, or are we off on a tangent?” Then, we’d work with James to implement. So it was a team effort, but it was just incremental experimentation and pretty ugly spreadsheet stuff on the back end before we could actually make it a system.

Sramana Mitra: How did you get these customers who were willing to work with you at this level?

RJ Talyor: Well, I mean, free trial means a lot, right? And a lot of these D-to-C brands are looking for people to help them grow. We’d go to trade shows or work with agencies. We actually created a tool that we called the Constellation Scorecard, which was a big winner for us. It is a benchmarking tool that gives you a playback of all of your data connected to our data co-op and where you rank in comparison to your competitors. That got us a lot of leads. Through trade shows, through cold outreach, and through the Constellation Scorecard, we’d typically start with a free trial, and we would prove ourselves into a paid engagement after that free trial.

Sramana Mitra: So you went after people and offered them free trials and that’s how you brought them in.

RJ Talyor: Yeah, which by the way, I wouldn’t recommend. I think that felt like our only play, but in one quarter, we did seventy-two free trials and converted one. That was terrible, so we actually moved away from free trials. Free trials was a good move at the time because we ended up getting a lot of data connections, which made our data model richer. However, we spent so many calories and so much effort to get performance metrics for customers who weren’t really that interested but were willing to do something free because it was no skin off their back.

So we found that free trial was a good way to build our dataset. However, to convert those customers into paid engagements, the only way to do it was through a paid trial or an annual agreement with a six month out or something like that. So, basically get the customer to get some skin in the game. If they aren’t willing to do it, then either they don’t believe in you or they don’t believe in the idea enough to want to pay a nominal amount. So, that was a good learning.

Sramana Mitra:  So I think what you’re underscoring is something really important. We hear this a lot. They say, “Oh, we’re going to offer the product for free and then we’ll figure out how to monetize later.”

People don’t understand how much work it is to service free customers. Resource wise, time wise, and as you said, calorie wise. It’s a very bad idea to offer anything for free without a path to monetization very, very early in the game.

RJ Talyor: I agree. It’s so much fun when you get a big brand or even a midsize brand who gives you attention and says, “Yeah, let’s do something. Let’s do a free trial.”

I mean, I think that there’s a place at the early stage, if you’re getting the value out of it, like if it’s truly a research activity. I don’t recommend it as a sales motion. I do recommend it as a research motion or a validation motion. Treating it as a go-to-market without some sort of automation, like a product-led growth approach is a path to burnout. At the end of that one quarter, we had spent so much energythat we had to pivot or change our go-to-market, which we did.

Sramana Mitra: Well, I think as a validation motion, it’s not a bad way to go, but you don’t want to implement things in a validation mode for free. You kind of want to discuss and understand the problem. You want to immerse yourselves in customers and really understand the problem. But I don’t recommend doing a lot of implementation work in validation mode for free.

RJ Talyor: Yeah, I think that’s smart advice.

This segment is part 3 in the series : Bringing a Generative AI Product to Market: RJ Talyor, Founder CEO, Backstroke and Pattern89
1 2 3 4 5 6 7