Metabase: How Ignoring Expert Advice Led to Real Growth – with Sameer Al-Sakran [439]

[00:00:00] Omer: Sameer, welcome to the show.

[00:00:01] Sameer: Thank you. It's great to be here.

[00:00:03] Omer: Do you have a favorite quote, something that inspires or motivates you, you can share with us?

[00:00:07] Sameer: Yeah, I was thinking a lot about that, and in all honesty, I think I'm just gonna quote myself just to start off on a great megalomaniacal slant.

But the thing that kind of kept bubbling through my mind as I was thinking about this conversation was the idea of playing your own chess board. To stick to your own board.

[00:00:24] Omer: It's gonna make a lot more sense once we start sharing the story. So tell us about Metabase. What does the product do?

Who is it for, and what's the main problem you're helping to solve?

[00:00:35] Sameer: Yeah, so Metabase is fundamentally a way for you to get the data that is inside a database. You have lying around your company. And expose it in either graphical form, visualize the data inside of it, run reports, generate numbers produce metrics, what have you and expose it to everyone in your company.

So we're intended to be a daily driver for the average person with the real job. Now we're open source. So our installer is typically very technical, but analytics is a multiplayer game and they're usually not playing it on behalf themselves. And so we are a tool that is used by anyone in the company.

It's installed. Because we're open source by developer. But eventually it is meant to be used by everyone.

[00:01:12] Omer: And I think there's about, I think, what, 70,000 odd companies that they're using meta base in some shape or form. And then you've got the, obviously you've got the open source side and then you're selling to enterprises is as well.

Give us a sense of the size of the business. Where are you in terms of revenue, size of team? That's tough.

[00:01:31] Sameer: We are just under a hundred people. I think we're gonna cross a hundred between the taping and when this airs. And we are well into eight figures at this point of ARR.

[00:01:41] Omer: Now, the business was founded in the, I guess the end of 2014.

And probably a couple of months after I started this podcast actually, so we can celebrate birthdays together. So what were you doing at the time and like where did the idea for this product come from?

[00:02:02] Sameer: Yeah, so at the time I was working at Expa which was a startup incubator and a venture studio.

We were starting companies. We were bringing founders from various companies in the valley, having them kick off projects. They eventually became funded. It became real companies. We were largely playing at kind of the seed to a range. And a lot of what I was doing there was building up shared infrastructure for the portfolio.

And so Metabase came out of a side project that I had where I was just like, didn't really like any of the commercial options for a host of reasons, but mainly due to just how heavy everything felt. And I wanted to create a lighter, slicker, easier version of a bi prompt. And it quickly became a viable real thing as opposed to nights and weekends. Project we staffed a team at Expa and then eventually spun it out in 2014. And so a lot of it came from me scratching my own itch of wanting something that I could just slap on top of the database that a company had lying around where?

There was no data analytics team, there was no data pipelines, there was no data stack. It was just whatever application you had at its database and it was essentially all about the agility of metrics or measurements or dashboarding.

[00:03:16] Omer: So I think when the whole idea of scratching your own itch, you, we, I think many of us are in a situation where you see a bunch of products out there and you're like.

There isn't something that does exactly what I want or this thing does this and that thing does this but not the way I want to do it. And if you're a developer, obviously life is a lot easier 'cause you can go and, build something and, be, live happily ever after.

But it's very different saying, I'm gonna build something for myself versus saying I'm gonna go and build something and turn it into a business. And as you and I were talking earlier, I. When it comes to these types of BI tools you were telling me there were like hundreds of them around at the time, so what was it that kind of pushed you to see this not just as something for yourself, but also a viable business opportunity?

[00:04:09] Sameer: I don't think I'm that unique or that special is maybe at the heart of it. And I think the things that I find irritating are the things that a large. Segment of the user or buying population finds irritating. And so I think a lot of it was just me seeing that. Most vendors had sacrificed usability at the altar of sellability, and there were just a lot of things that were being enforced by arbitrary like industry walls where I just wanted something that slept on top of Davis.

I didn't want a metric store, I didn't want a data cube. I didn't want, to set up any of these tools. I just want something like quick and dirty that would work. And structurally the industry wasn't really making that. And so the fundamental idea around commercializing it was just that. My problems were common, like nothing I was doing was exotic.

It was part of the tension I had where I felt like I had the most basic, simple version of this and there was no nothing out there that solved the simple version of five person company or five person team. At a larger company, we have an app with a database can I just see what's in there?

Like I don't want to, again, define metrics. I don't want to create a semantic layer. I don't wanna deal with all I just wanna see what's in there. And the level of difficulty of something to do something that simple felt out of whack. And so a lot of the notions of open sourcing it or commercializing it really stemmed from my perception of there being a fundamental, like missing piece in the ecosystem.

And and the other thing is that it was one of the joys of Metabase has been that we are targeting existing use cases. There's no, we haven't had to explain why someone needs a dashboard in 10 years. So it was very clear that we are doing something that was intrinsically commercializable and that we just had to do it.

We, in some differentiated or some better way. That there would be a, a light at the end of the tunnel with a business in it. So it wasn't like, sorry I'll jump back. But we were never manufacturing a demand. We weren't like transcending, we weren't building something funnily novel.

It was just a fairly basic desire that I had that I'm sure. A hundred million people I've had over the last year.

[00:06:26] Omer: How long did you work on it at Expo before you spun this

out?

[00:06:30] Sameer: I don't remember to be perfectly honest, but something on the order of six months. So like the desire was born very early.

But I wanna say that the actual code was not written until six or nine months in.

And how much money did you initially raise? What was the seed round like?

It was just 2 million. So fairly standard for the Valley at the time.

[00:06:53] Omer: Okay. Let's take a quick breather here. Are you struggling to bring your SaaS idea to life without a technical co-founder?

Gearheart is here to help. They're a product development studio founded by serial entrepreneurs acting as your fractional CTO. They've built over 70 successful products, including Smart Suite, which was named 2024 SaaS Startup of the year after selling his last company for $200 million. Smart Suite's founder turned to Gearheart to build his next big idea.

Their Ukrainian born team combines world class expertise with Silicon Valley connections, helping you scale your product faster and smarter. Gearheart's Leadership team even works directly with founders offering deep insight into startup success. Book your free one hour strategy session with their leadership team before the end of May @gearheart.io.

That's gearheart.Io. Now the idea with Metabase at the time was, a simple, lightweight tool. People can easily, download it, they can deploy it in five minutes or whatever. And so you're seeing this as a business opportunity. You start going down the open source route as a way to get it out there and then, and start to, just reach more people.

But. You didn't monetize for like you were telling me like for four years, so what, why, and what was going on in that time?

[00:08:13] Sameer: Yeah, so this is one of those things I'll freely stay. I don't know if that was the best of ideas. At the time, however, it was very intentional, which was we want to build a product that wins.

We're competing with a bunch of companies that know what they're doing. The options in the market are, I don't like them, but you'd have to be delusional to look at them and say that there's anything wrong with them. It's just I don't like the choices they made, but Tableau was a great product and remains a great product.

And so for us to build a viable business, the idea was we have to first build a viable product. And that took time. It took effort. It took a lot of iterations, and so there was our own desire to get adopted and then commercialized as opposed to concurrently adopted and commercialized. And at least in theory that's the open source playbook where you build the open source product, you put it out there, lots of people discover it, lots of people use it.

You build a community and then at some point you wave a magic wand and you build a business around that. And for us there is always this creeping like finish line of we've done enough to commercialize and it is very natural to just keep wanting to do more. I. There's always one more feature request we wanted before we felt comfortable.

We wanted to build one last thing, and it was, there was never a clear, once we check this box, we're gonna commercialize. But there was this subconscious insecurity and wanting to do more. And in, in retrospect, we probably should have shipped something we, we were confident selling earlier.

[00:09:43] Omer: Did you raise any other money in those four years?

[00:09:46] Sameer: Yeah, we raised the second seed round, two years in, so another two and a half, give or take.

[00:09:50] Omer: Got it. So that gets you through to the first four years. So tell me about the, how you went about the monetizing the product what had to happen for you to feel like you were finally ready to do that and then.

How easy or hard was it to get that first customer?

[00:10:13] Sameer: Yeah, I think we, it's funny because we, for the first couple years people kept trying to pay us and we just didn't take any money. So it's there was at least two deals that were, people were just basically hammering away. But we didn't really have something that we were comfortable supporting.

'cause one of them was an embedded deal where they wanted to embed us in their product and it was going to have a lot of distorting demands on what our fundamental product was. And we were always trying to build a mass market product, mass product, sorry, mass mass market tool as opposed to an enterprise tool where for us, success was always measured in tens of thousands of customers as opposed to a hundred.

And so we intentionally said no to a few people early on, but I think the first serious user that wanted to pay us, we eventually just couldn't do it because we needed to sign a 50 page MSA, we had the jurisdiction in some, in someplace in Europe, and we just couldn't do that. Like our lawyers just looked at it, they were like, you really shouldn't do this.

And so it was a case where we probably would've accepted their check. But the contracting was very onerous. And I think that this has been a long running challenge we've had where. Just the licensing of the license story we've had has been probably unnecessarily difficult, and I'm sure there's lots of unfor errors we've made there.

But back to the story, I think the big the big unlock was really just offering a way for people to embed charts in their products. We knew that we wanted to commercialize this for me, really day zero. Like from the very, very first versions, we knew that we wanted to charge money for this.

And then I think three or four years in, we finally let you go to Heroku application, give us a credit card, get a code that would remove the power by meta base and our little logo at the bottom of the chart. And that ended up just. Taking off, like quite a few people gave a bunch of strangers $300 a month on their credit card.

And at this point in time there were no salespeople, like literally no one to talk to. If you had a question, no one asked a question of, we didn't offer support. There was literally no support email anywhere to be found. And there was absolutely zero. Personalization or advertising or anything. It was just a CTLA deep in the admin panel where you would turn on a bedding offer on and off, and if you wanted to remove the logo, it would ask you to go to this website and bring back code.

And so we ended up getting nearly six figures in a RR from that. And then we probably proceeded to do something else.

[00:12:47] Omer: You start out like the first few years, if I just clicked right. You're focused on, we, this is open source. We're just gonna focus on building a great product and we are gonna monetize, but we're gonna do it when the time is right.

And then. When you did eventually do that, it wasn't like a lot of fanfare or going out and, trying to knock on a whole bunch of doors to make these big sales deals. It was like, let's enable it in the product. And interestingly enough, just people were able, there were enough people there who saw enough value to, to start paying you for it.

So I think he was saying like 70, 80k in or maybe close to a 100K.

[00:13:33] Sameer: I don't think it quite hit a 100K, but it was more than you would expect. And I think part of it was that we did have a pretty crisp idea of how we wanna monetize. So I think we were always very clear about licensing.

We're always very clear about where our like paid versus free line was. This was one of those cases where, we called our shot years ahead of time and it ended up being like the thing that we could monetize. But we did it out of order and a little too late. But it, the, the one thing that I think we got very correct, which was, the one thing that absolved of so many other bad decisions was just saying, if you want to embed this in your application, you will have to pay us at some point.

And so holding our ground on that and picking the right license and having the packaging of the product itself respect that and just not flinching. Because as people were asking for things as they wanted this feature, that feature, we held our ground on that line pretty strongly.

And I think that ended up being, again, the one good decision. They made up for a lot of mistakes.

[00:14:37] Omer: You basically have this self-serve model. You don't have a sales team, you don't offer support, and yet you get people paying and getting you on the way to the first six figures in ARR and so most people would then look at that and say, great, we've found something that's working.

Let's double down on that and I think you look back at that and intuitively you knew even at the time that was the right thing to keep doing. But then you got advice that wasn't the correct way to, to grow the business. So just talk about what happened, why did you change course and take a different approach.

[00:15:19] Sameer: Yeah, so I think there's what's been most interesting about our story has been the number of times where a very smart person would've looked at our situ, at our structure, at our setup, at our market, at our customers, and then advised us to do something where it ended not being the right call.

And, at the time and to be fair, this did work. It just, it's a question of ordering, but the commonly held wisdom at the time was you commercialize an open source project. By giving away the core and selling the manageability and the off, different off options permissioning, et cetera.

And so we'd ended up building an a quote unquote enterprise edition meta base. It ended up being a good idea to build, but I think the le the, the lesson that we didn't learn from the, just again, like the little CTA in the admin panel to remove one line and one link from our embeds we should have learned is lots of people are using it.

They'll find CTAs and we should be, I. We should stick to our guns there and just be like, Hey, here's a way to get this thing that you want in the moment when you're looking at it in the AM panel. And instead the generally held wisdom was you should talk to your customers. This is, we're gonna ask for a lot of money.

You should go and I. Establish trust. You should, guide them through the POC, you should learn about their use cases. You should be able to articulate value back to them, et cetera, et cetera. And so I ended up essentially just doing lots and lots of sales calls and eventually selling the first million in ARR or rather the next 900 will change by just respectively do having an inside sales operation.

Would just me doing it. And at that point we we hired some AEs and eventually grew the directly sold business a little more. But if I could do it all over again, it would've just been, hey, self-service was working really shockingly well, like far better than had any right to, and we should have leaned into that and stacked more behind it as opposed to saying, no, that's not what serious companies do. Do. Serious companies that have succeeded in this space or similar to space to ours are following this playbook. Let's follow that playbook. And to be fair, like lots of companies had succeeded and we were watching, Mongo and hashing and GitLab just do really well.

And that's what smart people did.

[00:17:36] Omer: Many founders would look at that and say, okay, regardless of which growth channel we focused on we've got traction. We are, we're, we're feeling like we, we've got some level of product markets fit. We've got the first seven millions, first seven figures in ARR.

Why do you feel like, was it a problem? And 'cause it, because then you had to you ended up shifting back to self-serve again, right?

[00:18:00] Sameer: Yeah. So to finish up a story is we got to give or take, I wanna say somewhere between 2 and 4 million in ARR. We ended up adding a cloud self-service option.

I, I'll have to like actually go through my like historical records figure exactly where, but at some point we built a cloud version and the cloud version started off with self-service and that blew up. And that just was on a tear from that point on and dwarfed the directly sold portion of business.

And so the, there were a lot of things that were all nibbling away at. At me and at us, and it's easy in the retrospect to be like, no, that was the path. But in the moment, there was a, the need to raise money. One of the biggest things was we were not cashflow neutral.

We did have to raise a Series A and every single investor that I ever met asked me the question that I still think is a dumb question. For what it's worth, which is like, what is your go-to-market strategy? And at the time I had a combination of a very sophisticated and a very naive answer. And that was, we're gonna get a bunch of installs, we're gonna put CTAs in them, people were gonna click on them, they're gonna give us money.

And this has the unfortunate set pair of characteristics, which is, it ended up being exactly right. And for our customers with our, our, at that moment, our DNA of the team, our skills, our abilities, the people we had access to. So the people that were actually engaged with us and the clear next ring of adoption, it was actually how things end up happening.

Now, unfortunately, this also sounds very naive and very realistic. And so this is one of those cases where. A sensible investor that has seen a lot of these stories go by would say, every tech founder, ever dreams of that and no one makes it work. It's really unlikely that you are gonna make this work.

Therefore, I want to see something that is a little closer to the patterns that have ended up working for other people. And so there's a bit of, there was pattern matching and from very smart people, they were very successful and they could very easily point to look, I saw this company, I'll do this.

And this included investors and ex operators and advisors. And there was just this general, common wisdom that self-service is very difficult. And most companies ended up trying to glue on a PLG motion after the fact. But it was actually fairly difficult to do this. And it's funny because in a way.

Our team was very unbalancing that's the kind of thing that we could do very well. Whereas like in-person selling was the kind of thing we're really bad at and we're really bad at building that, really bad at like organizationally setting that up. And so it was a place where. Any reasonable person would say, sight unseen.

There's a six person or a six to 10 person tech startup with no salespeople, no commercials, people. It's a bunch of designers, engineers. They want to commercialize an open source product by self service. And again, a reasonable person looking at be like let's get them some help. And the thing is in the rear view mirror, that is absolutely the right thing to say.

It just given our specific moment with specific set of tools we had, the specific amount of attention we had and our and the buyer that we understood. So we didn't understand all like the entire universe of possible buyers, but for the people that had come to us that had shown up, that had clicked on the website, that had download software, we had an affinity in for how they thought and for that slice of the world, we understood how they wanted to buy.

[00:21:39] Omer: I think the irony here is that if you hadn't been given that advice, you would've probably. Continued down the self-service path, maybe ended up in the same place, but I think it was also a matter of how much effort it was taking to get sales and grow the business. And this kind of, analogy of, whether you were pushing a rock uphill or whether it felt like you were pushing it downhill.

That's also I think a factor here. And so do you feel when you back that. Did it make sense that advice or were you deep down, did you feel like, this doesn't feel right to me or was that more of a hindsight thing?

[00:22:21] Sameer: I think I always saw the other path, but there was all, there was concurrently the awareness that I am a founder in my position, I am almost by obligation, delusional.

Like I, I need to see things that aren't there for this to make any sense. And that there are other people who are very smart and very accomplished, who I had a lot of respect for. They're looking from the outside and I don't think it's there. And so there's that combination of I see this path with a lot of smart people being like I don't know about that.

And there's also, it's my personal realization that there was a lot that I needed to learn. So at the time it was like, okay, I should learn how to do the sales thing. Like I should just do a bunch of calls. I should go through process. I should learn how to hire AEs. I should learn how to, hire VP of sales.

I, there are all these things that I need to learn and there's a lot of smart people that are willing to teach me. So let me, lemme go do, let me go do the work and lemme go learn and let me take in the playbooks of people that have made it work at other places. And I still think that a knowledgeable observer should, st.

Should have said what they said to me in that moment. Just by virtue of having being inside the, the container rather than outside. There was information and things that I saw that they didn't

[00:23:41] Omer: Let's talk about competitors and just positioning. We talked about the fact that it was, it's a very crowded market today.

It was a crowded market. When you were building the business in the early years, and how did you figure out how to differentiate yourself? I know you focus on saying building a great product, but what did that mean? And you, when we were talking earlier, you also said. I wrote down somewhere like you didn't spend much time looking at competitors, but you still regretted every second that you did spend time looking at competitors.

So why was that and how, how did you figure out your place in the market and how to differentiate yourselves?

[00:24:28] Sameer: Yeah, I think early on we had a very hilariously naive and simple growth strategy. And that was just win the taste test. So if us and two or three other options were being tested we wanted to be the first system standing and we wanted to be the thing that you wanted to use.

And the, that was the entire extent of our strategy. There was no content strategy, there was no social strategy, there's no paid acquisition. We weren't like trying to get it, onto anyone's radar in Nanos community, we weren't like worried about growing the awareness. All we did was say if us and Tableau and Superset and Looker are being used back to back, if there's a, if there's a stopwatch.

Who gets the system standing first and which one do you want to use and do you wanna keep using? And sometimes the answers to those questions were very flattering to us. The lots of things we had to work on, lots of things we had to build, lots of things had to fix. But we very intentionally.

Over indexed on that moment of decision where one person at two in the morning is trying a bunch of things out and we wanna make sure that their ability to give us a whirl was effortless and that they had a good time. So one of the things that we really chased in the early days was the statement that, I can't use this, but I really want to.

And in general, when we heard that was always cool, we did a great job. So there was no world where we could out Tableau. Tableau. There's, we weren't gonna be more enterprise ready than, at the time, MicroStrategy or Oracle BI or Cognos or whatever. We're not gonna be more enterprises than them.

So we're going for is we want to be the daily driver that you reach for first. And that was it. We focused on that, we just looped over and over again. We played with it, didn't like it, fixed it, watched someone else play with it. They didn't like it, we fixed it. And that was the extent of it.

And we com we built a webpage. I think we didn't change the webpage for years on end. And mostly just had a CTA, which was try or like it. And a lot of people tried it. Not everyone liked it. But enough people liked it to give us information to keep the wheel going.

[00:26:50] Omer: How are you getting feedback?

In order to do that, and if you're not if you're trying to ignore the competition, you're gonna focus a lot more time on how people are using the product, what they like, what they don't like talking to customers. I guess how much time were you spending on that and, you know how, what were you looking for?

'cause you're gonna get a lot of feedback. Some of it might be valid, some might just take you down, dead end. So how did you navigate that?

[00:27:22] Sameer: The answer is it was very complicated to navigate. So I wish there was some distillation, but in general, we did a lot of user interviews. We did, we would shop mocks to custom to users all the time.

Say, Hey, we're thinking about this feature. Do you like version A or version B? We would build prototypes to play around with 'em. We would rebuild an interface, try it, not like it, delete it. So there was just a lot of high conviction attempts. To fix certain key parts of the flow. I think the thing that is counterintuitive that I very much believe is we were first and foremost building something we liked using.

I think a lot of people over index on the danger of building something for yourself and the fear that just because you like it doesn't mean other people will. But I'd say that inverse of it is almost never true, which is I don't believe you can build something you don't like. Somehow magically strangers will like, so for us, we like really fought for that first like stage or that first milestone.

We like this until we like this. We're not, it's not ready. It's not no one else will either. And we really emphasize that. And we tried to keep in our minds the various classes of users, their motivations, what their data sets look like. We had a bunch of sample data sets and synthetic data sets.

We had real data from various places. We tried to see how it worked with various shapes and sizes of data. We looked at lots of different use cases and a lot of it was just. Us having to think through, all right, what's the consequence of this choice on all these classical users? And so it ended up being a very mushy puzzle.

It's not like there was three metrics that were optimizing, but we were trying to build a set of Legos that when used well together, could. It could represent a lot. It can be used to build lots of things. And I think we really abused the Lego set metaphor in the early days and still do, where we're not trying to build a bridge, we're trying to build a Lego kit, lets you build a bridge or a drag or a spaceship.

And the way you build that kind of product is very different than the way you build, I don't know, like a calendaring application or a CRM or something else. So to reiterate is we really indexed on creating something we liked first and then valid and doing a lot of validation on it with outside people.

[00:29:39] Omer: There's always a danger if you do that, that you're gonna end up building a whole bunch of stuff or just finding reasons to keep adding features or improving features and not necessarily, just putting it in front of customers. And so was that something you struggle with or. What did you have that awareness?

[00:30:02] Sameer: I think there was always a pathological need to ship, so I think that from the very, very early days we wanted to ship the question of will we ship? Always was a matter. There's, there was always the, is this ready? But fundamentally, we. We wanted people to see this. We wanted people to play with it.

And so once the open source version was released, it was just game on and there was a ton of feedback, a ton of complaints, a ton of, user stories, a ton of interactions. And so the pre shipping was torturous. Again, there's always the like, is it ready? How about this? How about that?

But I think once it's in the wild it just really shifts gears and you start having a need to improve what's already out there. So the bar is not, is it ready, the bar is, did we fix this thing that's really annoying me? Or that person, or did we add this thing that is really necessary instead of is the whole thing magically perfect now?

And so that transition from, getting something good enough to be a little embarrassing, but to be useful. And then once it's out there just saying, as long as, version 14 is a little bit than version 13. We're shipping it. So long as 15 is a little bit than 14, and that kind of fed on itself.

And so that never really became that big of a problem.

[00:31:21] Omer: Yeah no, that's a good distinction. 'cause I think that's, that situation that I was describing it is that pre shipping, right? You're, you haven't got the product out there and you're endlessly looking for reasons not to ship it because it's not ready.

But with you guys, the product was already there in the hands of customers and you are more in the mindset of. We are gonna keep shipping, we're gonna keep iterating until we like this product. And that's a different situation. I wanted, I wanna also talk about pricing. 'cause that was a really interesting thing that you guys went through.

Customers wanted to pay you in a certain way. Which you weren't willing to accept, I think either way where you told me. And so you ended up like trying a whole bunch of things, right? So what was going on there?

[00:32:08] Sameer: So early on we had this dream. The dream was everyone in your company should use Metabase and we want to do everything we can to get everyone in your company use my this, and that was considered victory.

We want to be used for everything by everyone. That we want to have our price pricing model be in alignment with that goal. And one of the beliefs that we held onto a little too long was that if we had per user pricing. There would be a, back pressure or negative feedback on adoption within the company and that, if something is just flat rate or is priced some other way, then obviously give everyone an account.

Why not? Whereas if it's, gonna cost you a lot of money per seat, you'll wanna hoard seats and you'll maybe want people reuse account or something like that. So early on we were on like installation pricing as a primary metric where like a number of users on one set of servers. We did per server pricing at one point.

We did like per dashboard or per database. I think at one point in time and throughout all of this I don't remember the exact customer conversations, but it's often some version like, can I just pay you per user or per creator or per reader? And we were like, no we want blank. And we don't do creator pricing 'cause everyone's a creator.

And in many ways we were just a little too principled for our own good. When we just started saying, look, what does everyone else charge? Okay, we're gonna charge that, or some version of that. Everything became easier. And even just copying the structure of what our users wanted to pay us simplified our lives a lot.

And so I'd characterize there's just being a downhill and we were insistent on not rolling downhill.

[00:33:58] Omer: So how long did that go on and what happened for you two finally? Except, per user was the way to go.

[00:34:07] Sameer: This was a good two years. So I'd say for a good two years we were just pricing in these weird ways.

And so they weren't weird, but it was generally not optimal for us or the customer. And I want to say that the main thing that happened was you start doing self-service and it was self-service. Having there be a transparent fair. Like price you could point to on webpage was a big deal. And that forced us to just come up with something that would be applicable to everyone, and it wouldn't require one-off negotiation.

Because there was no one right price or one right structure for all of our one-to-one negotiations. And so we ended up doing lots of things that made sense to that customer. And I think when we had to say, all right, we're gonna have like hundreds of thousands of customers on this thing, they're all gonna have an automated way to do billing.

We need to just pick a formula. And at that point it's like, all right, I guess the formula is X plus Y per user. That ended up working really well. Everyone was very comfortable with it. There were no further there's very little haggling from that point on, and we eventually took that into our normal negotiations as well, where just it was how people wanted to pay us and once we accepted that, everything became easy.

[00:35:18] Omer: Okay, great. So you started off in a self-serve route. You got distracted a little bit with the direct sales. Then when you went back to self-service or I guess your roots, that's when things really took off. And that was basically the main growth vehicle that got you from seven figures to, to eight figures.

And then also the pricing stuff we, we just talked about and some of the challenges you faced there. When you look back at that journey from seven figures to eight figures. What were some of the, scaling a business is never easy, but what were the, what was the biggest challenge for you to be able to scale effectively?

[00:36:02] Sameer: I think the biggest the biggest headache was going from there being, let's call it 15 people to 30. And that, but essentially going from a world where there was a handful of senior folks with people that report to them, to there being three layers. So I think that was maybe the transition that caused this most Harper.

And ironically this was exacerbated in the areas that we consider our strengths. So our ability, I think I mentioned our ability to delegate the things that we were just honestly bad at. It was very easy because lots and lots of people have a better way of doing things than you do, and it's very easy.

Just hand it over and be like, all right, you've got this. Let me know how it's gonna be. I think the places where we had done a really good job of organizing ourselves and having a system and having there be a bespoke, unique, interesting way of doing things we're the hardest to really scale up. So early on we had a very, like very tightly like threaded kind of combination of folks with product sense engineering skills, data backgrounds, and design sensibilities.

And it was very hard to untangle those. And now say you're the pm, you're the designer, you're the tech lead. And that originally there was just so much overlap and everyone wore so many hats. That we didn't fully understand how to do it in a way where those were different people with different responsibilities and we had to figure out what the communications team those were, how the process would work.

And looking back on it, it's really funny. Just the degree to which the things that we did know how to do were easy to scale and the things that we did know how to do ended up being the trickiest, the hardest to hire for, and ultimately. In many ways the most Harper inducing.

[00:37:58] Omer: That's really interesting.

'cause I've often,

[00:38:04] Sameer: when

[00:38:04] Omer: it comes to whether it's outsourcing or whatever, it always feels if you outsource something you're not good at, there may be people out there. I'm sure there are people out there who can do it better, but it, there's also a risk that you're outsourcing something that you don't fully understand how to do well, and.

And will somebody else, will you be able to manage that resource or whatever? But intuitively, I would've thought that the stuff that you do well would be easier to scale.

[00:38:33] Sameer: And I think part of it was that we weren't outsourcing, we were bringing in. We would hire, we would be very selective.

We'd hire the best people could find, and in general, it was very easy, relatively speaking, to find people that are way better at us than a number of these roles. And bring someone in, you trust 'em, they're on the team, they're in the family. They're gonna be around for a while and they'll build something that you look on and I'll be like, oh, that's actually pretty cool.

I could not have done that. And I think that was often the sense when people would show up and do again these functional areas where we just didn't have a lot of expertise or a lot of like interest or a lot of innate talent. And the places where we're like, no, I have like really strong opinions about how the product to design to engineering loop should look.

And finding people that were read into that system was very challenging. And finding people that would. That were at a fishing caliber where they want autonomy, they want to do their own thing. And there was always that tension of but like we figured this out. Let's just, let's not lose this.

[00:39:31] Omer: It's certainly been a interesting journey over those, I guess 10, 10, 11 years putting this when you started out and when you were like. When you were like, I'm not happy with what's out there. I wanna scratch my own itch. I wanna build this thing, which is simpler and just. Does the job in, in, the easiest way, did you ever envision it turning into what it is today?

[00:40:04] Sameer: Actually yes. And more. We literally drew out like projection lines on like the number of likes, number of stars, number of whatever. And what's hilarious about looking back on say, the seed fundraising decks is that in many ways we called our shots. It was like, this is how it's gonna go. It actually did go down that way.

And what's funny though is how much we messed up along the way. So looking back on it, I think that we were largely directionally. Okay. Like I don't think we made any big decisions. I look back on and, face pump, but I'd say there's just lots of things where I wish I could have done that a little faster.

I wish I could have not taken that misstep. And in terms of the scale, we were always playing for the scale. I think literally, I think there was a conversation before launch where if this thing doesn't get, 10,000 GitHub stars, it was all in vain. And it probably wasn't that lofty.

You probably used like a lot more colorful language, but it was some version of, if this thing doesn't get 10,000 stars, it's completely pointless. And so there was always this sense of we wanted scale and that a lot of us came from consumer web or consumer mobile companies that had just the way we just saw this huge adoption and we very much were programmed to see that as something both natural and something you fought for and something you could get.

[00:41:25] Omer: What I love about your story is that and hopefully the chess piece, the chessboard analogy makes sense. That you had a very straightforward, simple plan and you talked about, Hey, we're gonna have this open source product, we're gonna get a lot of people using it. We're gonna have this CTA, people are gonna pay us for this, and that's how we're gonna monetize it and build this business, and it's gonna be self-service.

And that I guess that was your chess board, right? That was the game plan and what you were gonna play out and people kept telling you needed to do different things and. And that they were well intentioned and probably as you said, the, the right things that you should have been hearing at that time.

But ultimately it was like you had a pretty simple strategy and plan and somehow it just, there was this, people just wanted, it just seems too simple.

[00:42:20] Sameer: I think part of it is that. In that era. Also, I would say that the standard naive, delusional dream of a technical founder was that you'd build a self-service business.

That was almost the stereotypical trap to avoid, which is, there was a number of, press articles about this. A number of people going on record saying this doesn't happen. This is not the way the world works. And there's also the honest maybe uncertainty of if we had fully played in the, on the other playbooks, we would have been further along because I'd say there's lots of ways we could have shortened the path.

I think there's a lot of years we probably could have speed run or skipped. And it's more than possible that if we had. Been better wired for, the more standard playbook, the success would've been larger or faster or different. So it's hard to rerun history, but I do think that I.

When I look back on the path that we took and I think about where could I like, cut years off and have turned a 10 year journey into a four year journey or five year journey. It's just interesting where it's oh I wish I could have snipped this year out and just thrown it away, which I could have snipped this year out and just thrown it away.

Because like in that year we worked we tried, we experimented, we did things, but it was like we were just like on the wrong path. Like we shoulda just. It's done the other thing and you really just snip that out and then put it towards the end. Because I do think that, for example, right now we are finding success with direct sales and that is something that was gonna be part of our next chapter.

It's just a question of ordering and, understanding what your unfair advantages and applying your unfair advantage over and over again, and not trying to pick fights that you're not gonna win.

[00:44:04] Omer: That's great. And I think on that note, let's wrap up. Let's go into the lightning round. I've got seven quick fire questions for you.

What's one of the best pieces of business advice you've received?

[00:44:12] Sameer: Just charge money.

[00:44:13] Omer: What book would you recommend to our audience and why?

[00:44:16] Sameer: I'm drawing a blank on that one. So for the decline in Fall of the Roman Empire by, but for reasons that just are tedious. There you go.

[00:44:24] Omer: What's one attribute or characteristic in your mind of a successful founder?

[00:44:28] Sameer: Irrational perseverance.

[00:44:30] Omer: What's your favorite personal productivity tool or habit?

[00:44:34] Sameer: An empty notepad that you just fill in every morning.

[00:44:37] Omer: You're like, I think you're the. Some, I was talking to somebody yesterday about this as well, just obsessive about taking notes. It's it's inspired me to get my notebooks out.

What's a new or crazy business idea you'd love to pursue if you had the time?

[00:44:50] Sameer: There's a zillion people doing this, but personalized language tutoring using LLMs. Nice.

[00:44:58] Omer: What's an interesting or fun fact about you that most people don't know?

[00:45:02] Sameer: I really enjoy being corrected.

[00:45:04] Omer: You enjoy being corrected.

And finally, what's one of your most important passions outside of your work?

[00:45:12] Sameer: Some form of movement practice. So it's changed over the years, but some way just be in motion. And stay in motion.

[00:45:19] Omer: Nice. Thank you for joining me. It's been a pleasure unpacking the last 10 or 11 years as a story. Lots of really interesting insights about building a business like this. And I appreciate you sharing those lessons with our audience. If people wanna check out Metabase, they can go to metabase.com and if folks want to get in touch with you, what's the best way for them to do that?

[00:45:45] Sameer: Just send me an email, sameer[at]metabase. And yeah, it's been great to you on here. Thank you for having me.

[00:45:50] Omer: My pleasure.

Thanks man. And I wish you and the team the best of success.