A Product Market Fit Show | Startup Podcast for Founders
00:08:36 · Chris Slowe
[00:08:36] S0: I think he did a pretty good job on the first on the first batch. I think the best way to describe the actual experience was the shared context of – there was no other room you could be in than that room where everyone was literally going through the same thing that summer. Everyone had just established a company. We were all learning how to build basically web apps live at a point when web two was just becoming a thing that was being talked about. No one knew how to build scaled web products or at least there wasn't a kit to do it. You're learning live and then being able to once a week at dinner, have a chat with some other people who were doing the same thing and being like, oh, I literally just solved that problem this week, here's what I did; first discoveries of some of the new things that come along with CSS, which to this day remains one of the worst toolkits imaginable. I'm sorry, I went back in engineering mostly by training. I got out of that early. I mostly have completely uninformed opinions on it at this point, which still are funny. From my standpoint, we were working on, sounds really mundane but desktop search. Me and my co-founder at the time were a couple of physicists who just couldn't find PDFs on our fricking hard drives that we had downloaded because the search toolkit that existed on a computer was pretty terrible at that point. We ended up trying to solve the general problem with search on desktop. This is about, I don't know, I think it was that fall that Mac OS shipped an update where they actually gave you the search bar. That's how far back in the past it was. I think we were competing with that stupid dog in Windows. It would appear and like, oh, you look like you're trying to run a search. It's like, yeah, this is completely useless.
00:10:18 · Pablo Srugo, Chris Slowe
[00:10:18] S1: My life changed when I stopped clicking on files and I just open up the search and pull up a document. [00:10:24] S0: Yeah, it's like command, space, bar. Everything is command, space, bar at this point. Yeah, that was still a relatively new new idea. Of course, now at this point files locally really? Everything's in a browser window and it's probably stored somewhere else that you just have to find it. [00:10:39] S1: I guess my question is, when you joined, what was the MVP of Reddit? I mean, in Reddit it was actually almost known for not changing or at least not changing all that much, and so how similar was the first version of Reddit to what we see today?
00:10:54 · Chris Slowe
[00:10:54] S0: I think a lot. A lot of the basics existed all the way back to that year. I'd say for that summer and for the first version of Reddit, it really was just a front page with a bunch of links. You could submit content, you could give it a title, and you could vote on it, and that was it. There was no conversations. There was no discussions. If I remember right, the first comments appeared later that year in December of '05, give or take, after I had started. It was a major addition. I remember also really clearly setting the tone for Reddit's culture literally one of the first comments that appeared was something like, "Reddit has comments now. I guess it's now just slash dot. We should all move on to the next thing." It was great while it lasted. It's like, oh yeah. I mean it's like eternal September all the way down. Voting plus comments goes back to 2005, basically. I think the next major edition we did was when we added subreddits closer to 2007, 2008. I would say that was one of those things that actually slowly started appearing where initially we had created a set of – we just started finding that the amount of content being submitted was so large, we had to have some an organizational principle around it and the original idea, we started breaking down topics. We curated in the first crack about 10 or so, now we call them communities but at the time we just called them subreddits because they're a chunk of Reddit around a given topic and that was things like science, politics, basically clear delineations that the community wanted to separate out.
00:12:24 · Pablo Srugo
[00:12:24] S1: When I look at Reddit, at least I think the most powerful asset you have is this incredibly strong community, and more so than maybe any other social network frankly is the defining feature to the point that communities now subreddits can move markets. It's pretty incredible. How did you get that going? Everybody wants to build a community, but it's so hard to do. I'm just curious in those early days, I remember stories of seeding the threads and stuff, putting things there to get conversation started, but do you remember how you guys got that going in 2005, 2006 because it's easy once there's content but it's hard when it's a blank page?
00:12:55 · Chris Slowe
[00:12:55] S0: Yeah, 2005 was – I remember really clear this is right before I started where for that first summer, all of the content that was on Reddit was either being submitted by Steve and Alexis or it was being submitted by a bunch of YC and associated because we were the first users as a group. There was a point that Steve likes to talk about where he finally took a day off on the weekend. He just decided I'm taking Saturday off and then he went back to Reddit. I mean, I'm going to butcher his story but he can yell at me later. He comes back to it the next morning and he is like, "It's going to be stale. I'm going to go submit it. I have to dig up a bunch of content and post it." He opened up Reddit and there were 25 fresh links on the front page and he had nothing to do with any of them and it was just like this like, oh, the system seems to be working. From that point on there's no need to start creating this – that particular community, that very first community, which was just reddit.com, started to run itself and it just started becoming this place where interesting content would just appear and bubble up. I think one of the things we got right really early on was the underlying math of just how content became popular on Reddit and how long it stayed up there was very simple. It was optimized for being able to index it on statically so you could actually render really fast. We had it publicly in the open source version of Reddit from about 2008 till 2017. Honestly, that algorithm didn't change markedly for most of the first 15 years of Reddit. It was a very straightforward setup just based on age of the content and the number of votes it got. A lot of the work went into just making sure that the votes that we had good anti cheating logic to make sure people weren't doing things that were inorganic. The core of it was very well established. Later on, that same process worked for communities where we noticed up front that there were a bunch of topical areas that were starting to form, and so starting to peel those off of a front page made sense. There wasn't any real work to seed them because the content was already present.
00:14:44 · Chris Slowe, Pablo Srugo
The real change was in 2008 when we actually opened up the floodgates and just said like, "Hey, if you think you have a good idea for community, go make one." Then that was the initial burst of just all sorts of random stuff that users started coming up with. I mean, that's where actually AMAs, which are a cornerstone of Reddit right now, were primarily something that formed on its own. IAmA was formed around then. Most of the major communities that came later were ones that were just organically created by users and that's actually the best result. [00:15:21] S1: That was almost like adding fuel to the file. There was already momentum and you're just feeding and feeding and feeding it.