00:26:58 · Guillaume Moubeche
[00:26:58] S1: And I think you talked about, like, the the big swings at Descript. I think you you did something that generated, you know, like, a lot of the the revenue with... Can you please, like, share a bit the the growth loop that that you put in place there?
00:27:16 · Sandy Diao
[00:27:16] S0: Yeah. Descript is a, again, an audio video editing product. A lot of the users there are gonna be content creators, whether that's an individual content creator or a business content creator. And if you really think about the product without referencing any data, if you really try to have intuition about it, you'll figure out that the product itself is really meant for users who have their own built audiences. So this idea is essentially that every user that you acquire, that content they produce is meant to be seen by many, many other people, which means that in theory, there's this opportunity or this network effect, a user to user network effect, where one user can always equal n equals, you know, more than one user. So in order to make that true, we need to figure out how do you tap into the distribution that your users already have. And so one of the pieces of intuition I had here was we can essentially figure out how to get content creators to sell Descript on our behalf. But how do we do that without me reaching out to every individual content creator setting up some kind of content or partnership agreement there. And so one of the experiments I ran, even despite not having a lot of firm data around this being a growth lever for us, was setting up an affiliate program. And the affiliate program was very simple. We used a third party tool. We didn't build anything special or in house. And we essentially used this tool where people could sign up for a tracking link. That tracking link would measure the conversion of a Stripe event, meaning that somebody paid, and they would get a, you know, a percentage or a kickback on the referred paid subscriptions that they drove to our business. And we would just send this out to new users who signed up. In our onboarding emails, you know, not even that openly, but in the footer, it'd say, here's the affiliate program. You can go to our website. There'll be affiliate program. And when we first launched, we obviously made an announcement email, allowed for us to get our first, base of affiliates. And we launched it, and we said, let's keep the cost minimal, the build minimal, and see what happens. And it turns out a lot of people were interested. We opened the program to everyone at the time, so anyone and everyone could become an affiliate. We learned a lot of lessons along the way.
00:29:20 · Sandy Diao
Some affiliates are really bad. They do things like bid on your brand terms through Google Ads, which is terrible, and we eventually banned that, of course. But then there were also really great affiliates. As soon as we launched it, we saw organic content just pop. Right? If you went on YouTube, you saw a lot of tutorial content. You saw a lot of videos. We thought this is great. When you actually looked at the results and the data, you saw that they actually did refer paying customers because they're so... Creators understand the product. Their job of content creation, when we talk about the product, it comes across very authentically. And so by putting the link to the product there, even the creators themselves don't feel like they're running too much from an ad per se because they're running these workflow videos. So that ultimately netted us in realizing that content creators themselves, whether through affiliates or even eventually paid placements through influencer marketing, was gonna be a massive channel for us. And that eventually, you know, drove, you know, upwards to a third of the the business for, quite a number of years. And that's something that would not have come about if we were to have diligently looked at our dashboards, our traffic sources, and try to, you know, be overly analytical about what was going to work. It just really came from this understanding of who our users were and this intuition about the space itself.
00:30:39 · Guillaume Moubeche
[00:30:39] S1: That is really cool. And to your point with referral system, actually, when I first launched the referral system, same using, like, one of the most famous platform out there, I actually, like, I went on a vacation, and I typed like the name of our company online, and then I saw an ads, and I wasn't doing actually like any Google Ads at the time. So I was like, how come... So I reached out to someone in marketing, and I was like, when did we start like doing ads? I thought we were not doing this thing, no, no, no, no, we're not. And then I looked at it, and I could see like the, you know, in the URL, the ref code of the affiliates, and when I went back, you know, to the affiliate program, I had seen that in the last six months, whenever my head of marketing was reporting to me that the ref road program was working really well, that it was basically linked to our growth, etcetera, I was super excited, and realized that one partner was making the most of his money by just like buying our name on AdWords, and just, yeah. So in the end, stopped it, but again, I think when it's done in the right way, you can definitely get like some amazing results. And I think you wrote on your blog, that is actually, it's a really cool blog, like, about finding your niche in AI. Like, how how... What's what's different versus what happened in the past, and how would you frame it, you know, like, for people trying to find a niche with with this new technology?
00:32:17 · Sandy Diao, Guillaume Moubeche
[00:32:17] S0: Yeah. And are you referring more to finding your niches in the right type of product or feature to build, or more so how to use AI in your existing workflows? [00:32:26] S1: I think, like, whenever you you want to find a niche, you know, for for your product and... Or in the new features that you want to develop.
00:32:34 · Sandy Diao
[00:32:34] S0: Mhmm. Yeah. I think a lot of the startups I've advised in the last few years, especially when they're building AI native products from scratch, one of their biggest challenges is trying to figure out if they go very, very, like, narrow and horizontal and go after a very, very specific persona. And that tends to be the easiest in terms of shaking up the product road map, or they go very horizontal. So they try to build something that has, you know, unlimited use cases. Right? And they're like, you know, sky's the limit with AI tools and features, and we just want users to be able to be creative and do whatever they want. It's always, you know, this this very tricky balance. Because on one hand, if you build something that's very narrow, you firstly get called a wrapper. And then secondarily, becomes very hard to remain competitive, especially as, you know, more of the general purpose models can essentially tackle many of and probably all of the tasks that exist out there today. And then on the other hand, if you build the really, really broad product, the challenge is you're not gonna be better for the most part than foundational models, and you have to essentially figure out how to win that market. So it's this very hard question to answer. I I will say this about, you know, sort of figuring all of that out. I I think that what I've learned is it feels that TAM is virtually infinite here. I don't mean that in the sense that the number of people on the planet is skyrocketing or growing. In fact, there's a little bit of a reversion on that number. But instead, I mean more so that people have two distinct behaviors that I think are worth tapping into. One of them is that switching costs are relatively low from tool to tool. Meaning that if you are building better experience, the better design experience, the brand that people trust more, you can still win in the grand scheme of things. And then I think the second thing to note is when you're trying to sell the product itself, people are willing to use more than one tool for the problems that they're solving. So it's this idea that even though you see maybe a foundation model doing something really well, you can actually still get a mark there if you can solve it better for them in like one corner of the universe or one specific area.