A Product Market Fit Show | Startup Podcast for Founders
00:41:08 · Chris Slowe
[00:41:08] S0: I think by about 2015, Hipmunk was in a good steady state. From my standpoint, a lot of the technical problems were solved or were just a matter of playing an existing playbook. There wasn't a lot of difficult work to be done there. What happened for me was I ended up consulting for Reddit mid '15, because at that point they were starting, starting to scale up a little bit more. The company was up to around 70 employees. They were looking at actually building a mobile app. They were looking at doing something that – some of the stuff that I had originally specialized in Retrospect was I became an infrastructure generalist. I became probably one of the first SREs in the team just as nature of the beast. I also ended up being one of the first people to work on what amounts to being – we eventually called anti ego, which is our user safety stuff. I was secondary to another guy, Mike Aldi, who is a good friend of mine now, who was the first – I don't know. At the time he was more the troll poker. He would help to fend off the trolls and put them back under their bridges and got really good at building out the toolkit for making sure the users actually had some amount of safety. Coming back the things that were top of the list for me were working on that safety toolkit, and so I ended up doing some consulting along those lines; talking about mobile usage; talking actually about data science, which wasn't really a function that existed much at Reddit that point, which again, my title at Hipmunk was actually chief scientist, which most of my titles historically was a joke but at the same time actually it was accurate. I was a scientist but also data science had come on the scene around that time, and so it made sense to say, well, Chris can work in analytics. Then by the end of that year, it was – I've made this analogy a bunch of times, but I have to repeat it. It's like it had the same feeling as dating your ex-girlfriend where it's like you dated a long time ago and you broke up for very good reason, but that was like five years ago.
00:43:06 · Pablo Srugo, Chris Slowe
[00:43:06] S1: Have you changed? [00:43:07] S0: You both changed as people, so let's just see what happens. That was actually exactly how it felt where I had thought that I had emotionally separated myself from Reddit and then actually by about midyear in '15 Reddit was going through some tough times. They had a major blackout that was infamous, the first real modern blackout, and there was this possibility that Reddit would cease to exist and that hurt in a way that I wasn't expecting. It was almost like, oh God – I'm twisting metaphors in a bad way here but I'm going to do it. It was like seeing your kid in tough times. We worked really hard to get you going and then you find out that later on they're leading a rough life, so maybe it's time to swoop back in and see if you can actually help. [00:43:52] S1: Especially when it was a known brand, real brand with real users, a real community, the thought that that could go away, I could see that being not easy to accept, not easy to swallow.
00:43:52 · Pablo Srugo, Chris Slowe
[00:43:52] S1: Especially when it was a known brand, real brand with real users, a real community, the thought that that could go away, I could see that being not easy to accept, not easy to swallow. [00:44:05] S0: Yeah, and I think to some extent too, we just had a different model for how to approach Reddit, the community as well of – I remember historically there's always been pitchforks and torches on Reddit. That's the nature of the product. Prior to 2010, some of the ways to solve the problem would be the community would get ahead of Steve about something and then Steve would appear and just Spez would just comment, no. I remember there was one case where it happened in 2008, 2009, where he just replied no, and then the response he got was, "Okay, guys, let's just call it clearance from the advents. They're not into it, so let's just move on." [00:44:39] S1: It's crazy the persona of Reddit. It's pretty wild. It's just pretty wild to look from the outside. It's so unique. [00:44:51] S0: It's very much so [00:44:52] S1: I was going to ask, you mentioned the mobile app, what took so long? App Store comes out in '08, '09. I just have to ask that. It's nothing to do with product market fit but you would think –
00:45:05 · Chris Slowe
[00:45:05] S0: I know. I mean, that's a fantastic question. Actually, we had a couple of early forays into it. I think that I can start off with saying the first answer is honestly capacity. We had a very small team at that point. Like I said, we had four or five engineers tops. Rather than spending time learning a new new kit, we initially decided to just build like a mobile experience that was actually not bad. Early on there was actually – I wonder if it still works. I'm going to really quickly just check and see if this thing is still in existence. No, it's not anymore. Okay, good. We used to have a particular view of Reddit that was optimized for mobile browsers. At this point 2008, we weren't just targeting iPhones, we were targeting Blackberry, Research in Motion and Palm OS and all those features where we didn't know what was going to be the clear winner, so we just picked that. Second, we actually did at one point bring in a contract firm to build us a first version of a Reddit app. Then subsequently after about 2008 when we open sourced, we started having more of an ecosystem of third party usage of Reddit appearing and early on there was a whole set of third party apps that developed. In fact to the point where I believe in 2013, 2014 it's after I left, but Reddit bought one of the major iOS apps called Alien Blue and then relied on that as the iOS solution for a while with no equivalent solution for Android. By about 2016, the idea was that we were looking to probably do a bit of the look of Reddit and do some cleanup, and at the same time it was worth actually investing in some first party clients that really were taken from the ground up that had a particular more modern style.
00:46:52 · Pablo Srugo
[00:46:52] S1: The other thing I wanted to touch on is the monetization piece. I mean, you mentioned it wasn't really a focus I guess until about you guys came back in 2015, 2016. I'm curious just what worked? Fundamentally, what were some of the big things that you did? I mean, you think, okay, you're going to make money on ads, but what were some of the key things that you implemented that really drove the real revenue that you have now?
00:47:13 · Chris Slowe
[00:47:13] S0: Okay, well, two things. One of them is going to sound like a cop out but I'm going to say it anyway. One of the things that actually had the most impact for revenue was just fricking grit. The reason I say grit is because trying to bootstrap an ads business – it really was bootstrapping at that point because we're talking about very low numbers even in 2015. Trying to bootstrap an ads product on top of a "social media platform" was extremely difficult given the fact that we had made a bunch of different choices from the rest of social media. We had no real user data. We have basically interest data. Everything is in the public and in the clear. You go to the front page, you're going to see the good and the bad of Reddit based on popularity. That's the value proposition of the product is that it's going to be an unvarnished view of what people are talking about. No wild gardens, no fancy look and feel. It really was educating advertisers over a very long time and really kudos to our ads team and our sales team then to now of just having to continue that path of saying like, "No, we're differentiated. We have a different value proposition than what you got off the social media. We're not really social media but we'll pretend to be, or at least we'll call ourselves that." It's the cleanest bucket. The second thing that actually came up early on we actually did do a self-serve product in 2008. The first version was my build where we wanted to actually have users be able to basically buy screen real estate on Reddit directly with the credit card. From about 2016 on, it was rebuilding that into a modern auction, actually doing it in a CPC model as opposed to a CPM model which was a lot harder to explain, building out things like tracking pixels; really building the whole kit and caboodle of what is expected on a modern ads platform but again, with a start date of 2016 as opposed to everyone else who had started in the early 18s. Lots of catch up, but also the nice thing about catch up is that you do get a chance to take advantage of the modern toolkit. We made it all the way into the 55 minute mark or so without me saying LOMs. This is a tech podcast. Somebody has to say LOM at some point. I think it's on the screen.