← Previous · All Episodes · Next →
PMF is one pivot away with Ant Wilson from Supabase Episode 43

PMF is one pivot away with Ant Wilson from Supabase

· 37:19

|

Ant Wilson 00:00
Product Market Fit is always just a pivot away. And you can go from having like one user sign up per week to like 10,000 users signing up per week.

Jack Bridger 00:13
Hi, everyone, you're listening to scaling deciles, the show that investigates how Delta's go from zero to one. I'm joined today by ants from Super Bass. And this is a really exciting day, because I'm a big fan of Super Bass. I use it on my projects, and really excited to have come along today. And thanks so much for joining. Thanks for having me. And could you tell us a little bit about Super Bass and about yourself?

Ant Wilson 00:39
Definitely. Yeah. So super bass is the open source Firebase alternative. And it's fully open source, so you can host it yourself. And it comes preloaded with auth, which you can think of as like, an alternative to auto zero, real time, so you can stream changes to your database, file storage, so you can store images, videos, anything you need. And it is just Postgres under the hood. So it also comes with all of the amazing Postgres extensions that have been developed over the last few decades as well. And for me, I'm originally from Liverpool. I'm the founder and CTO of Super Bass. This is my third startup. And I've spent probably about 10 years just building hundreds and hundreds of failed prototypes and until until we eventually got some, some love in the open, open source community with with Super Bass.

Jack Bridger 01:43
Yeah, and that kind of leads us on quite nicely to the first question, which was I saw that you posted the other day. That was it in 2014, you read this book, launch pad in an airport? And then that completely changed the direction of your, your life?

Unknown Speaker 02:00
Yeah, it's quite, there's a funny backstory here. So when I was a kid, I rarely wanted to be an archaeologist, you know, when people would ask you what you want to be when you grow. Archaeologists dig for dinosaurs and pyramids and stuff. And then, but then, you know, there's no one in your life. There's no one around you, who is an archaeologist or has become an archaeologist. So it kind of just like, died away. And then I found computers and got really into computers. Fast Forwards. I went to study for Masters in London, and a guy I was living with was studying Egyptology. And he was going on digs in Egypt, like literally digging of mummies. And I was like, wow, why didn't anyone tell me that this was a realistic, like career path, I was so disappointed that no one like in my life was like, Oh, this is what you want to do. Like, you can go into it, there's a path The path to becoming an archaeologist. So then, when I, I always just think in terms then of like, you know, you need these people in your life who are like role models and can show you this like realistic path to the thing you want thing you want to achieve. And then so since I was in uni, like studying computer science, I kind of had this dream then of starting startups. But again, like I didn't know anyone personally, who just started their own company and had this success. So you end up just then getting a job and working for other people. Until I picked up this book in the airports, which is the launch pads. It's a zoom in on one of the Y Combinator cohorts and just basically describe what happens in the cohort. And it just gave me that like, clear path of like, oh, this is how you can like start companies. And actually, all you need is like a few friends and access to the internet. Like it reinforced the idea that like, you don't need a lot to just start building. Well, dev tools to start building your own startup. And I just got so much inspiration from the stories in this book, that it gave me what I needed to then just go in to defend and you know, quit my job, start trying to get involved in startups build my own. I kind of just started me on that path.

Jack Bridger 04:27
Yeah, that's amazing. So it's like, Yeah, lucky that there wasn't a book about Egyptology. And so you have like a really like, kind of unique way of doing things that super based and it seems like that kind of came from a lot of those experiences that you had and very seemed very formed by yC as well. Could you tell us a bit about how you operate Super Bass?

Unknown Speaker 04:52
Yeah, I think like going back to the start was was very interesting because what One important thing I learned from from that book, and from just, you know, reading PGS essays and all of this stuff is that like, product market fit is always just a pivot away. Like, it always feels like it's kind of, it's there somewhere, it's just out of reach, it's just around the next corner. And, and one thing I learned from those stories is that you can go from like, having basically nothing, you know, having five users signed up to your service. And, and then you can make some small change in either the product itself or to go to markets or the audience that you're trying to address. And that can just change everything. And you can go from having like, one user sign up per week, to like 10,000 users signing up per week, and kind of with these tweaks, and that's kind of why I'm involved a lot now in in marketing and super bass as well. And it's always thinking about like, Alright, you've we've got this product, which we think was amazing. But like, how do we talk about it? Who do we talk about it to? You know, it's kind of like, it's just the database at the end of the day, it could be used by really any type of developer. But who are we really going after first and why? And just thinking those things through? So that's some of the lessons I guess I learned from ultimately taking part in Y Combinator and all of the literature there.

Jack Bridger 06:32
Yeah, it's, it's really interesting. You say that that like, it's so that one, one user a week is just one little tweak away to becoming the 10,000 users? Could you tell us a bit about like, maybe like, what it was like before you found that tweak? And then if anything changed afterwards, in how you kind of did things?

Unknown Speaker 06:53
Yeah, definitely. So before we found that tweak. And we were, well, we were marketing ourselves as real time Postgres. So we were big fans of Firebase, we use Firebase and our previous startups. And we were aware of some of the limitations and some of the things that people maybe didn't like about Firebase. And one thing is that it's proprietary. And so we said, you know, we think Postgres is an amazing database. But what are the bits that are missing? And the first bit that Kapil, my co founder built was the real time engine on top, so you could stream the changes into your front end. And so we started marketing as a real time Postgres. But that's such a, that's such a niche audience, compared to when we kind of on a whim change the tagline to be open source Firebase alternative, but with the same product, basically. And it was like day and night, it was like flip flicking on that switch that literally took us from eight hosted databases on the platform to 800 in three days. So like someone, one of our early users, like, shared the website, again on Hacker News, with this new tagline open source Firebase alternative. And it just clearly like resonated with a much broader audience of developers. Unsalted does a better job of communicating rarely what this is supposed to be. And, and the difference then is, when you've got eight users, you're sort of like, you think you've got this amazing product and platform and your, like, people should love this, but your confidence is a bit shaken. Because you're like, oh, like, what are we doing wrong? To then when you have all these, like hundreds of people signing up, you then have the confidence to be like, Oh, this is awesome, this is great. And it should also do this. And we should also build this thing. And you know, and then once you really like, get that momentum is when it rarely gets fun, you know, you can start like you know, we built the auth product on top. And we built the file storage product on top. And you can really just started running with it. And to be honest, since since that moment, it's also been just trying to keep up with the with the growth of the platform and you know, it's open source so we just kept like tons and tons of issues created every day and just trying to stay on top of it is is most of the most of the work.

Jack Bridger 09:32
Yeah, actually could I could ask this I think asked for a mirror similar question. But like how, when you have such a kind of influx of like, requests and tweets and all this sorts of stuff, like how do you like filter through and like prioritize, and then also your own direction and vision.

Unknown Speaker 09:50
It's hard, man. It's hard because the bias is always towards like the last thing that you have, right? It's just naturally like, you see You tweet someone complaining about, oh, you know, I can't log in, I want to log in with discord on oath, whatever, that just naturally becomes your highest priority because you've just seen it. So we've spent, we've put a lot of work into this like feedback mechanism. And it really is like a fire hose. So we got all the GitHub issues, GitHub discussions, there's a feedback widgets on the websites, and support emails, Twitter, like all of this, and it all automatically gets funneled into notion. And then it gets paid to the relevant teams. So if it's all related tickets, it ends up on the auth teams task board, and then they just get to triage it. And then they decide whether it's, you know, we've got, we got 50 people asking for this, or is it just a one off? And if it's a one off, maybe we'll keep it on the side until we get more requests? And so we've put this pipeline together, it's still a challenge, just because there's just so much to triage. But at least it's a little bit better than just reacting to the most recent request.

Jack Bridger 11:08
Yeah, I can, I can imagine how hard that must be. Yeah, that's really cool. And what happens if it's like a kind of general like, requests, requests, it doesn't like fit with, like a team.

Unknown Speaker 11:21
So so it all goes into like a master database and notion, and then it sort of gets tagged by the team. So it's still like viewable for everyone. And it also allows us then if it's, if it's relevant for two or more teams, we just add multiple types so that everyone sees it. And yeah, actually notions been pretty good. We've been able to leverage it for quite a lot of internal processes where I guess previously, you would have needed like, specialized tools. And but yeah, it's been good to us. Yeah.

Jack Bridger 11:58
That's really cool. So you talked about like, kind of open source Firebase. And in my last job, we were using Firebase as a mobile app. And I remember hearing about Super Bass like ages ago, and must have been really early. And it was like, it does resonate. Because when you're using Firebase, you're always thinking like, Okay, this is not gonna scale forever, it's gonna get really expensive, starts to not be the best tool. And I think Super Bass is like, trying to be like, the best tool at the beginning and the best tool, like all the way through to the end. And that must also create some challenges, because those users at the beginning have very different to the ones at the end. And I wonder, like, how you're kind of thinking about balancing those kind of needs? And

Unknown Speaker 12:42
yeah, definitely. I think like, the people we saw most recently who smashed this out the PA was for sale, right? So for sell came in. And for the first time, well, not for the first time. But I think in terms of dx, it was the most impressive way to go from I'm just on my local machine, to now I have this like globally distributed front end. And the key for me always in my head is like, if I end up on the front page of Hacker News, is my app on my website gonna stay off, like that's the test, because you can't lock in us and then you just straight away get whatever 10,000 hits. So Well, that's the first test is like the Hacker News test. And for WsL, it's just you just know, you're not gonna have to worry. And, you know, it might cost you a little bit more. But you know, it's going to stand up at least. And so that was the first barrier that we sort of met with, with like Firebase was, you know, the certain technical limitations because of this, no sequel implementation. And, and also like, this isn't just for for five years, but also for a few of the different cloud providers. Is this like, surprise bill? thing that everyone's terrified of, right. So if I do end up on the front page of Hacker News, am I gonna get like a five grand bill at the end of the month. And so addressing those needs, at the start of super bass like saying, like, these are things that we definitely care about that we want to be that we want to make pass the product, and was very important. On the scalability side, like my background actually is in large scale storage systems. So from the start, we, the backend scaling just came a bit more naturally. So it's not something we had to like hire specifically for, you know, when we were a small team of like three or four people. And a lot of the scaling problems were just from like, we would max out the number of sub domains. We were allowed on AWS, or on Cloudflare or whatever, and that they were the things that would actually break in the app when we would have large amounts of Traffic was just like our limits on various cloud providers. And, but for the, for the users themselves, like, I think the knowledge that it is just Postgres and Postgres can absolutely scale to millions of users. And is one thing we definitely want to communicate to people straightaway. And that's why like, Postgres is kind of prevalent in our marketing as well. Because it is, it gives developers that level of comfort, right?

Jack Bridger 15:33
Yeah, I think it's something you've done a really good job on. And like, it's really clever how you kind of overcome that. At the beginning, no one wants to trust a start up, but then you're kind of like, it will, you know, don't trust us trust trust Postgres. And that's, like, you know, it's really, it's really cool how you did that, I think,

Unknown Speaker 15:51
yeah. And what's really cool about Postgres, as well as, like, they, the pace of the postgres development team is also extremely fast. And so we find ourselves, like having to deploy a new version of Postgres basically, every year. Because, and we actually, we hire a couple of people to just contribute to Postgres core. And because we also want to, that's one of the long term goals is just to make Postgres itself more scalable, and make it more scalable for everyone, not just super bass. And so yeah, so just going back to the question, that combination of like dx and, and scalability is just always been our like, the two main thing of the bowling alley like the two main Yeah, not a very good analogy.

Jack Bridger 16:45
I can't remember. Yeah, that's, I don't really read it. I don't you read? Okay, and then one question that everyone kind of talks about, I've had multiple sources tell me that you are the Twitter, the meme Lord on Twitter? Could you confirm or deny this? remark?

Unknown Speaker 17:06
I won't confirm or deny. But I will accept that I'm extremely passionate about the Super Bass memes. No, it's funny, because it's, it is one of our most successful marketing strategies. And it's so funny because like, basically, someone did an analysis, actually an incredible analysis of our Twitter. And they broke down the whole strategy without even speaking to us. And, you know, they basically nailed it got it spot on. And then they shared it to Hacker News. And then the Hacker News comments started coming in. And people were like, Yeah, I love super basic grades, but I just wish they wouldn't post these stupid memes. Like, I don't know why they share these memes, like, and I just wanted to, like, write in the replies, it's because it works. You know, like, me, like me was a great, like, you know, share them with friends, whatever. But we wouldn't be sharing them on the super nice Twitter if they just went so damn effective, honestly. And, and I think we're in a little bit of a unique position in the in the dev tools market, where like I said, super bass can be used by any type of developer, you can be on PHP, me on Python up on JAMstack. And it's, it's applicable to all basically all, all developers. And so we can just as part of our marketing strategy, market, develop a culture and go for brand awareness that way. So we don't need to be so targeted, where we only talk about the benefits of Super Bass, we can just basically say, you know, Oh, isn't this a funny? Isn't this funny about Junior junior developers or senior developers or whatever. And and then we get brand awareness with some developers somewhere. And it's turned out to be quite effective because then taken this like impressions driven like marketing strategy, and means that then all of the search traffic we get is basically branded. It's people are searching for Supervisors, Supervisor or supervisors. How do I do this super base, blah, blah, blah. And so we don't have some of the same like marketing problems have to do with like SEO? We haven't we haven't had to dive deep into like, SEO optimization, because we just get so much branded traffic at the moment.

Jack Bridger 19:33
Yeah, that's that's really a good point. Yeah. I think it was Jacob that did the the report on Twitter. Yeah, he's, he's been on the podcast as wise. Brilliant. Brilliant. Yeah, it's and how did you like get to this point of, was it like a kind of trial and error thing and we'd like you just

Unknown Speaker 19:54
the advice I give to any well, not just Dev Tools founders or founders in general. is like, you have to try every channel. And you have to try every strategy to know what works before we started superbooth We weren't Twitter users. We didn't understand the platform. We just tried it out. And we were like, Oh, like this actually works. And then it's like a case of all right? What are the different styles of tweets? What are other people successful with, you know, like a little bit of memes, a little bit of like developer centric questions, a little bit of pushing the product, and finding that blend over time. And equally going to Reddit, and, you know, testing out the strategy for different subreddits it's all about and we've tried all kinds we've tried, you know, Instagram, Tik Tok in person meetups, and just land and the, you know, the efficiency basically, of each channel. We just kind of discovered what, what worked and what didn't?

Jack Bridger 20:58
Do you like look at kind of just general like awareness? Or do you kind of try and think about, like, kind of people creating a database? Or like,

Unknown Speaker 21:07
is there like, I think it's like, you definitely need to be aware of the insertion point. So when, when, what, exactly, at what point in someone's day, or action? Do you want them to think of Super Bass unfroze? It's when they're starting a new project, right? Whether it's in wear, or whether it's on the weekend, it's a side project, it's like, I want to try out some new technologies this week, like, what am I going to try, and we need to make sure we're in in the brain at that point. Or so we do keep that in mind. But I think we take on more of like, just trying to be everywhere. And I think Swix did this tweet a few months ago, which which really solidified it for me, which was thinking about spaced repetition. So like, just memory in general, is you need to get your brand in front of someone on the Monday on the Tuesday, on the Wednesday on the Friday, and then it doesn't matter whether it's like one day is Twitter, Tuesday's Raddatz, you know, Friday's a podcast, or whatever. It's just like, you've got to cram as many. Just get your brand in front of these people. And also make sure you're delivering the best message as well. So like you said, about like open source Firebase and, you know, building a weekend skeleton millions. It's like, that's just the message that's going everywhere all the time. And maybe like, if we're doing a specific product launch around off the then we'll, you know, that will be a slightly different message. But then we'll make sure that that's the message that gets pumped out in, you know, in the blog on the on the partnership announcements on whatever. But it's just like, yeah, just trying to be everywhere.

Jack Bridger 22:58
Yeah, it's it. It makes sense. Because I guess it's really hard to like, insert yourself, as you said, at that point, otherwise, unless it's just like in someone's brain to think of Super Bass for starting a project.

Unknown Speaker 23:12
Yeah, luckily, I'm not the only developer who gets distracted by Twitter every 20 minutes. So that's also why it's a good place. Because, you know, you're being inserted there every every hour in the day. Yeah.

Jack Bridger 23:25
Yeah. Yeah, that makes sense. And one thing I was thinking about when you were saying this is like, obviously, there's like when people are starting a new project, but it must also be like the like, considering switching are like getting frustrated with like, does that play into it at all?

Unknown Speaker 23:41
It does. The thing with databases is they're very sticky. Like migrating your database, for for lots of reasons is a challenge. And something we've had success where there's people migrating off. So if it's from Octa or zero to Super Bass, it's a lot easier, we can just basically like, import all of the users into the super base instance. And offer just a generic database workload, it's quite difficult, actually. It's actually one of the things that makes us defensible, as we grow, as well as that we know. It's harder to move away. Although again, like with the branding and positioning, it is just Postgres. And we let people know that. If you do want to leave our platform at any moment, you just do a PG dump, and then restore it in any other Postgres database, and you're good to go. So that's kind of a little like marketing edge for us. But to answer your question, it's way easier to get people on boarded for new projects than it is for migrations.

Jack Bridger 24:51
So it's like you're more you more want to kind of like nurture the next like huge projects from from day one.

Unknown Speaker 24:58
Yeah, and you know, Even within large enterprise prizes, most of them have these internal development labs as well for new products. So you can definitely get in with new products in larger businesses. And we've had a bunch of success with that. And just, you just reminded me to go back to the meme discussion earlier. And another reason it works actually, is because and this is why people think memes are stupid as they think I want to sell to the enterprise. And therefore I need to be really serious. But like, that's not the case anymore. You know, the developers who work in a large company of 10,000 people are just like me and you, the, you know, the guru on the internet, they think memes are hilarious, like, and that's who we're selling to, we're not doing like takedown through the boardroom trying to sell to the CTO, it's all bottom up. And therefore, like, that's why it works. Because they are the same people who are trying to build a startup on the weekends, or they've got these, you know, the side projects going on. And I don't know whether it if it was ever different, if maybe people that worked, worked at IBM were more serious or not. But I think you know, these enterprise devs are really just like everyone else.

Jack Bridger 26:22
Yeah, that's such a good point. Yeah. Very interesting. Yeah. So they're not just these robot, very serious people. That's

Unknown Speaker 26:32
exactly, exactly. So again, it's kind of like, it simplifies the messaging. Because when we say, you know, you can get started with a full Postgres instance, in 60 seconds. And that applies for both someone start on a side project on the weekend, and someone starting a new project within the larger organization. And actually, that's often the biggest selling point, because they are way more resource conscious. Like, if you're starting a side project that weekend, you accept the fact that I'm probably going to spend 200 hours or 500 hours or 1000 hours over the next few weeks working on this. And organizations have to justify that cost. And so it's, it's, it's a big selling point for us. Yeah.

Jack Bridger 27:20
Yeah, that's really exciting. And that kind of like ties in, I think, to the final question that I have some questions coming from Twitter. But as kind of a super base user, what what is the future? And we also got, I should say, we also got this question from Jonathan work. So what is what is the future of Super Bass?

Unknown Speaker 27:45
Yeah. So actually, there's a I don't know if you've ever seen the Facebook, red book, The Little Red Book, the little culture book that they put out years and years ago, you can probably find it somewhere on Twitter. And I think Amjad from Rappler, tweeted a couple of months ago. But one of the pages I really liked was, they said, you know, we think about strategy over 30 years. And like, where do we want to be in the long term, and then we use that to plan the next six months. And, and that's kind of the way supervised operates as well. So are like, we don't have a very fixed roadmap for like 12 months, 18 months, 24 months, that's kind of flexible. We want to make sure that every three months, we're reassessing you know, what, basically what people are asking for, but we have a really clear picture of where we want to be. And so while I can promise is, we're gonna keep doing these large weeks, every three months, and every single product is going to receive a facelift. So everything is just constantly on the go, intense development, everything from auth, to storage, to the database to Environment Management. And I also just don't want to give away the big launches, we've got planned for launch week seven in the first week of April. This is this is me dodging the question.

Jack Bridger 29:13
Okay. Keep the surprise. Yeah. Yeah. Okay, cool. So that's really the focus is like, when you say the further years like, is that some, like big vision for super bass that you kind of have?

Unknown Speaker 29:27
Yeah, I mean, we always talk about like, what's the what's the end states database look like, you know, what, what's, what's the capabilities of like, the ultimate database and it's some blend of, you know, like, it should be able to do your analytic analytical workflows and your transactional workflows. And you should be able to store like, You're, like, manipulate your data within the database without having to pull it all out and involve these other systems. So there's a lot So things that we want it to be, I think the general waves that we're riding at the moment is just like Postgres itself, like I said, it's just under such intense development types. So they're emerging as, as the favorite database at the moment. And, and also, this migration from much larger companies have moving off the likes of Oracle. Like, they've realized that they don't need to spend 10 million a year on some, like Oracle fleet, they can actually just run Postgres, I think Oracle are, like challenging this by also rolling out Postgres. But it just reinforces the trends of like, that's where the industry is, is moving towards, and that's where we want to be. And so if we could someday be like, the kind of open source oracle of the world, that would be a good place to be, I think,

View episode details


Creators and Guests


Subscribe

Listen to Scaling DevTools using one of many popular podcasting apps or directories.

Apple Podcasts Spotify Overcast Pocket Casts Amazon Music YouTube
← Previous · All Episodes · Next →