#005

April 13, 2021

Scaling Through Experimentation

We had Seth Bindernagel and Aaron Glazer on this episode to discuss their process of helping grow plenty of incredible companies: Firefox, Twitter, Strava, GrubHub, and more.

The host

Nima Gardideh

President of Pearmill, ex-Head of Product at Taplytics, ex-Head of Mobile at Frank & Oak. YC fellow.

Our guest(s)

Seth Bindernagel

VP Marketing, SoloLearn. Ex-Twitter, Strava, Firefox.

Aaron Glazer

CEO of Taplytics

About this episode

Most modern growth teams use a culture of experimentation as the main process to grow.

There are lots of unknowns around how this process is meant to be run: what tooling to use, how often to run tests, and how to prioritize.

We're excited to have Seth and Aaron on our podcast to discuss their process of helping grow plenty of incredible companies: Firefox, Twitter, Strava, GrubHub, and more. We'll cover what sorts of growth people to hire at different stages of your product, how to think about product-market-fit after the early stages of the company, and how to think about growth infrastructure (tech, marketing, ops).

The Hypergrowth Experience

Get notified every time when we release a new episode.

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.

Transcript

Nima Gardideh: Hi, this is The Hypergrowth Experience and I'm your host, Nima Gardideh. 

I'm the co-founder and CTO of Pearmill. And we help startups grow through art and technology. Most modern growth teams use a culture of experimentation as the main process to grow. There are lots of unknowns around how this process meant to be run, what tooling to use, how often to run tests and how to prioritize. We're excited to have Aaron Glazer, co-founder and CEO of Taplytics and Seth Bindernagel, VP Marketing of Solo Learn and ex Senior Director of Growth at Strava. 

On this episode, we discussed their process behind helping grow plenty of incredible companies like Firefox, Twitter, Grubhub. 

We cover what sorts of growth people to hire at different stages of your product? How to think about product market fit after the early stages of the company and how to think about growth, infrastructure, tech, marketing, and operate. Aaron and Seth are both fascinating to listen to hope you enjoy the conversation.

Aaron Glazer: Hi everyone. I'm Aaron I'm co-founder and CEO is happy clinics where the number one feature management and experimentation platform out there to help you increase your core metrics whichever ones they maybe 

Seth Bindernagel: And I am Seth Bindernagel. I'm currently the VP of marketing at Solo Learn. We are an EdTech platform that teaches people how to code. We are in high growth mode, so probably facing a lot of the issues that many of you who have dialed into this call are facing in working in tech for the past 15 years on and off growth teams at various technology companies like Mozilla, Twitter, et cetera. 

Transcript of the episode

Nima Gardideh: Yeah So one of the first things I want us to sort of tackle and this is something I've been trying to figure out as well since I've been mostly in the startup world and it's been hard to sort of see the composition of the growth teams and how they evolve and what they look like mainly because I've always been at the ground zero sort of companies instead of being in the later stage companies.

Although we now deal with some of that with our clients I'd love to get your thoughts on How do you think about the different types of people you need to [00:04:00] hire You know think about it Let's say you know seed to series a series a to B and C and then later on when you're a scale-up how do you think about the types of people you want to bring on You know I I'll let you structure this yourself but I'd love to hear your thoughts and we can sort of dig into it efforts 

Seth Bindernagel: Aaron, you want to kick it off? I mean, I obviously have some opinions, but why don't you get us started?

Aaron Glazer: Sure. I think that you know, I'll frame the conversation in saying that there's no, there's no perfect roadmap to this. And it's different for every company, every startup and every situation it's, and it changes constantly as your business evolves. evolution of your business is, is in some ways going to, be in front of the changes in the types of people.

And then for some roles is actually going to be behind it. But in general, I think when you think about growth, if it was easy, somebody would have written it down and everyone would, would be doing it. So there's this big element of number one, are you looking for people? And I think across all of the different stages, it's probably fairly universal, but looking for people [00:05:00] with this high degree of intellectual curiosity, people who can see the world and ask questions about why is something that is.

To then inform what you're doing on your end. But Seth, why don't you hop into it? Like you can from the different stages 

Seth Bindernagel: I've noticed that the earliest stages, you kind of need the people who um, have an entire portfolio of skills. And so. Those are people who have to be able to do analysis. Oftentimes they need to know sequel. They need to be able to write queries. They need to be able to write copy. They need to be able to comment on design.

They need to be able to write user stories. They effectively need to be like product leads almost. Because they're going to be asked to do everything in those earliest stages. And this I'm typically talking about pre-seed companies, seed companies um, those that are scaling companies that haven't really raised a lot of money because they're still establishing product market fit.[00:06:00]

And then, and they're starting to dip their toe into growth. So it's a lot of times it might even be a little too early to start growth, but finding that really quality person who can do everything is really key hire. And then as you scale, you know, you're starting to raise into your like series a and series B.

You're getting to a point where. You actually are having inside the teams that are focusing on specific things and you can start to specialize. And I think at that specialization point. You start to like branch out. Now my case, I recently engaged Nima and his team because we're a series, a company going into series B.

We don't have a fully fledged like growth team. We're not ready to in, in house paid marketing. it's not a core competency that we could ever do on our own right now that lines up with the scale we're trying to achieve. So Nima and team represent like a perfect opportunity for us to outsource an entire growth team effectively.

At some point, Three years now for who knows what it'll be we are going to hit a scale where like Nima and his team are either [00:07:00] not going to be able to achieve the same performance. They will have achieved what they wanted to for us, or we can't afford them anymore because we just get so big and we're going to have to internalize it. Like it happens in every stage you get So big that you have to start making these determinations to bring people in. 

Nima Gardideh: It's for sure I'm going to be the last point. You just going to be able to afford us anymore at some point. [laughing] 

Seth Bindernagel: Yeah. But you know, it's, it's happened to me on two separate occasions where our teams have grown so big that we started to basically just hire internally because it was no longer cost effective or whatever it may be. And that's that's okay. Like that's a totally acceptable thing. And so it's really, I think depends on where you are in your company.

If you are an early stage company pre-seed or seed, and you're still thinking about product market fit, you may not want to go full head on into, in fact, we could say, I don't think you want to go head on, into like building out a whole growth team right now, because you don't even know if you have a product that fits with the market is demanding that moment. You get that product market fit. That first hire to me kind of feels like they have to be that jack of all trades that can really [00:08:00] run hot. 

Nima Gardideh: Isn't it hard to find those people like correct me if I'm wrong. And I think Aaron you you struggle with this too When you guys were starting your own growth sort of functions in house isn't it hard to find someone that has all these like things like they're usually way too expensive to be coming in An early stage startup that only has like one or $2 million to work with and I know maybe all of us here are like that where we we can write our own SQL queries and do design a little bit and even code a little bit and and understand marketing very well But those types of people and correct me if I'm wrong you guys pull you have hired more people than I have They're expensive Like they're hard to find And usually they founders you know because they if there's that this good maybe you should go find your own company uh and work on it. So what do you guys do there? I think Like that's an interesting problem. 

Aaron Glazer: Yeah. I mean, I think that you need to set the bar properly. The cost of the person is usually usually the limiting factor. Um, I think it's more about the availability and finding somebody in that kind of role. And I agree with you. I think a lot of the people who are doing that are probably off founding [00:09:00] their own things because they can, that being said, if you're able to find somebody who shares some of those attributes, you know, they can tick some of those boxes and most importantly, see the other ones as an opportunity.

So the fact that they don't know SQL, if they're looking at that and saying like, I don't know, SQL, and this is going to be amazing. I get to learn SQL. Then they'll learn SQL and they'll accomplish what you need to do, or they'll find someone else to do it. The person to look out for is the one that says, I don't know, SQL, and I don't know how we're going to do this because you don't know at an early stage, you're not going to know how to do anything.

Like you're not going to know how to position your product in the market. You're not gonna know what channels, like there's way more things you don't taking that on head first is so so key. 

Seth Bindernagel: Yeah. And another one is, I'm a person who may say, like, I don't, I don't know SQL, and I'm not sure you should be hiring people that need to learn how to do SQL. Like can't you just like use a tool. That'll do that for you. You know, people who are like, Less eager, I think, and are looking for tooling solutions to solve problems.

Also very strategic [00:10:00] thing. Like, yes, you should do that. But I think you wanted, like Aaron said, you want to tap into the people who are eager and it, yeah. It's like unicorn hunting. You're you're never going to find the exact person.But Nima, to answer your question, I think the other thing too, is that there are a lot of people in Silicon valley who know they are not founders, but they know they're early stage employees and they've made a lot of money and they want to jump back into the next rocket ship.

And so a lot of it is like, how well can you recruit these. You know, how much can you sell yourself as a person, as a team member and the idea that you're building to get them on board with the future promise of excitement. You know, there's a lot of that. I think when people have made their money and they they're sitting on some Facebook stock and they're happy and they're, maybe they're like recently married or whatever, they're looking for that next challenge.

And they want to jump into something that they feel is more meaningful to them because there's a decent likelihood. They've spent the past four years grinding at Facebook and learned a ton of stuff. [00:11:00] But being like, Ugh, I want to go do something that's like small and exciting and new and can get big and I want to be in on the ground floor.

So that can also be a way to recruit people where the promise for cash upfront. Isn't there. Maybe you, maybe you recruit them with equity or whatever it may be. I think there's a lot of that happening too in Silicon Valley or anywhere, anywhere else.

Aaron Glazer: And then those people start teaching the people around them Right. So those people end up making the entire team better and you can really learn a lot from them when you can't find somebody who shared like you can no you can't five of those 

Seth Bindernagel: people.

But I think the other thing too, it's sort of a quick side, we're going to, we're going to diverge here really quickly. I want to hear your guys' opinion about this Nima. Maybe, maybe you have an opinion on this, but Aaron, I'm sure you probably do too. Sometimes when you bring folks like that in, I think one of the things is like, if you're the founder or the manager, it's really important to keep them like thinking about the project that they're working on and not always go back to the, when I was at Instagram, we did this, this and this cause I've also been at companies before as well, [00:12:00] where like, people will always kind of reference their, like what we did in the past.

And I think one of the things I've learned being at like three or four high growth brands now is that there's some underlying principles that are universal and then every brand is. And like, that's another thing to be aware of when you're looking for those early people. Are they willing to accept that they know a lot and at the same time may not know anything at all.

Do they have that like level of humility built in where they're like, I don't know, let's get in there and find it out, you know? Cause if they have that, that's like a really strong marker. 

Aaron Glazer: And if they can synthesize like you know when we when I was over here this is what we did And like I think that the reason it worked is this And you know can you apply that principle and this new situation because you know if it worked at Ben then other people are going to know it and it's not going to work the same way today 

Nima Gardideh: I've been on the receiving end of that a bunch of times especially with you know we work with some of these like ex apple or Facebook or Google folks that are VPs at these startups now And and they'll say stuff like that And it usually like makes me roll my [00:13:00] eyes because yeah it was great because you have Facebook behind you when you were doing this it was very trying to scale a tiny startup and it's a different problem Or maybe they misunderstood the power of the brand the company that they were with before where they're like well it was so easy for us to get this these costs dynamics out of out of these networks and over the reasons because he had a global brand and behind you when you were when you were trying to convert people and it's just a different different game yeah I totally understand that a bit maybe the underlying principles thing is an interesting one to delve a little bit dig in What are the underlying principles that you think are you've seen a couple of times in the you you've clearly been part of now a few of these big companies Firefox Strava Twitter what are the things that you you carry with you through these 

Seth Bindernagel: I think people who are willing to, um, explore the entirety of the funnel and understand attribution, as an incredibly complex problem is like a universal. that I, that I like having conversations [00:14:00] about. again, it doesn't, it doesn't mean like if someone came in to like, well, when we did this, this is what we saw.

So therefore it's going to work at your company. that's a no-no, but like, if they're willing to get in and like take my understanding of like the current state of attribution to the next level, or push me on something that I might have deep understanding of, so that I can then make myself better.

So, like, when I got to Strava, I was I was really just getting into, um, being. Deep deep into like leading like a highly technical growth And one of the people that came in, um, that we recruited in, really has a deep, great understanding of things like attribution and incrementality and things like that in inside the paid world.

Now, admittedly, I, I didn't know much about incrementality, you know, I'd like read about it and done, you know, read some blog posts or whatever, but like we were doing paid at such a huge scale that we really wanted to see if our ads were incremental and his ability to like, Not [00:15:00] suffer from like the curse of knowledge and be able to explain to me as in like, almost like layman's terms to get me up to speed.

So then I could start challenging him on what we were doing was great, because like, there were moments where like, I was like, I don't know how you would incrementality. Like, and then we would go back and forth on it. The concepts were understood, but like, then you get into those tactics. So I think like for me, if you're in the world of growth, like attribution a, um, table stakes.

Like if you can't how to attribute your users to the various tactics you're putting into place, then you're like probably not going to be a good fit for a growth team.

Nima Gardideh: Yeah that's a really good 

Aaron Glazer: one.

I throw another one in there Um you know I haven't been at those companies but when you interact with the folks who were there and you know Seth I'd love to hear your view from the inside but you really get the sense of the people love the product Like there's almost this magic fit between the person and the product where you have to love the product You have to really [00:16:00] understand it You've probably spoken You are a user you you interact with them and you really understand what's starting to make the users tick to then ask those questions of like well why to drive at the tactics 

Seth Bindernagel: Yeah, that's a really good point. I think I've been lucky to work at companies where I've loved the product and have been a user of the product. Now, Mozilla Firefox early on. Great. I was one of 450 million people the only browser out there outside of IIE until Chrome came along. Ooh. Yeah. I love that product. So did everybody else in the world? Twitter was kind of the same thing, like early days. Strava. Might've been the very first product where that really was the case where I have been an avid runner all my life. And I actually actually ran in cross country and track in college.

So running has just been a thing. And certainly when mobile devices came out and you started to be able to run with your mobile phone and be able to track your, looks, your location, it all of a sudden like became a thing that all the running community adopted. And when I got into the [00:17:00] Strava and it was like every single person in that company was like an outdoor athlete and.

People would be at work all day long and like sweat clothes and they'd show up dripping and sweat. And they'd work at their computer all day long and sure. They get a shower. It wasn't a disgusting place, but like you know what I mean? Like they were like a very active members who really love the product and solo learns the same way.

Everyone that's on staff, it's still learn. Um, Really loves learning. They love, they love learning science. They love, they love diving deep into what are the things that makes the human brain retain information better when you learn, when you're trying to learn a subject and they love code, it's a bunch of developers. So, um, it's fun to work at those companies. And you can, you can get an interview on that as you're kind of suggesting know, you can try to find people who are like, yeah, man, I really love, teaching people stuff or I'm a teacher or whatever it may be.

Nima Gardideh: it's just like a equivalent of almost like founder marketing But almost like employee or pocket or market fit [00:18:00] where w working on you're just very into we talk about this a lot We have the same same thing like think agencies you have to be very into cracking rows for people and it's it's if you're not excited about this like conceptual thing you're going to hate it we found out this like the hard way I think through many years and now we look for this like sort of gear for mastery on on what you're doing those people they just love it and they are way better than me almost at it because my thing is just company building I'm much more interested in into that than than figuring out if the specific market is going to crack the people that we work with are like that So this this definitely resonates and I think also makes it so that everyone enjoys their work more which is a huge part of how culture is built inside companies. 

Seth Bindernagel: And by the way, I think eliminates people in the B2B world from being able to like recruit on the same thing. Like it is possible to find people who are passionate about technology and passionate about what the customer that the underlying B2B technology serves, which can be another proxy that [00:19:00] there are great. I mean, Aaron, maybe, maybe you experienced this yourself. It's just like, if I woke up and said, man, I really love developer tools. I can't stop dreaming about developer tools. Like, no, you probably don't but do you love helping technologists build awesome things? Yeah, I do like that, you know? 

Nima Gardideh: Definitely not that loud developer tools. [laughing] I know people who are obsessed with building B2B, SaaS companies So like I don't know I think they exist all forms of people exist and definitely there. 

Aaron Glazer: But I think that totally exists Like you know for me I'm constantly in awe of the companies that we get to work with and the people that we get to work with And you know for us we get to be part of this story and this journey from some of the most amazing companies that we ended up working with And I'm sure Nima is probably the same for you Like you know you get to be a part of the story and a part of a journey And like you know you're helping to build that that end product that is so powerful through the teams that you work with even if you aren't on that team directly like you're a big contributor to that success which I think is is one of the allures on the B2B, SaaS side. 

Nima Gardideh: [00:20:00] Totally Yeah And I think like what's what's interesting on B2B is that the problem you're solving is a lot more clear You're like okay I'm helping you get faster answers from your data For instance you know if you're building some analytics tool or I'm helping you be more organized and building some productivity tool it's just it's very clear compared to maybe some consumer apps where you're like Hey this is a social product you know you sometimes using it you know it's it's at least like the folks that I know are obsessed with B2B is because they see the value immediately And also like you're immediately charging for the thing quite often And it feels good because you're building something and immediately people are getting value out of it You can have one-on-one interactions with them where they telling you I love what the software looks like or how it functions And it makes my life easier at work I think people enjoy the sort of feedback that they get A type of feedback is a lot more personal 

Aaron Glazer: Yeah I think a lot of times like you know it's thought of as B2B and there's this disconnection like this disconnect from the consumer but it's really like we think of ourselves as a B2B to C [00:21:00] company you know and that connection from what we do to the end consumer like can never get lost because like that is like actually the magic that makes us work And the fact that we're working with another party on a B2B side it's still like it is all for the consumer at the end of the day which I think ends up making so powerful.

Seth Bindernagel: I was only going to say it was a re-emphasize what you said that B2B oftentimes had such that crisp definition customer, which adds to that excitement because it's right there for you. 

Nima Gardideh: I want to touch on the attribution stuff because I think it relates to another topic we were discussing in the pre-interview which is how do you approach infrastructure when you're trying to build a growth team Do you go out there and by segment or by Taplytics to do your experimentation with feature flagging Do you build that in house You know I think maybe Aaron you can start here I'm sure Because you probably hear all these different reasons why people don't want to buy your software [laughing] So I'm curious to how you guys think about it. 

Aaron Glazer: Yeah One of the moments that stands out for me was early days I think it was sort of near to when we [00:22:00] came out of Y Combinator and we started selling our product we got this big pitch in a really big brand name company and we're all sitting around the table and everyone was like super into it and they're all excited And then there was uh the classic hippo like your highest paid person's opinion came into the room and said like Hey no no no We don't we should just be building everything in house Like we should be doing everything Like what are we doing Bringing this in And and there was this junior engineer in the room and the junior engineers I was like Hey our line of business is this like building the best customer experience software. And just like all of these other services that we need to build on top of our software that's someone else's they should be doing it. Like we we should be building the fastest transactions so that this runs smoothly.

And I think once teams get that focus of like what is your core focus and what should you be really putting your energy in And then what's been solved by others and they're doing it at such scale that it's better than you could do [00:23:00] In-house and it costs less on an ongoing basis. And a lot of the times when you think about it like oh let's do it in house versus Let's bring some other product and whatever it is you know this isn't just for us it's really easy to look at the upfront cost of it and say like oh well you know it's this amount per month And you know we've got this engineer sitting right here Let's let's get them to do it But then that system is going to keep running and running and running and then the person might leave and who's going to maintain it who even knows how it got there And I think a lot of those things once you've been burned a few times on it then you you you sort of develop that memory of let's let's build the things that we're great at and we want to support forever and then let's buy the things that we can get off the shelf that uh support the use case 

Seth Bindernagel: I would, I would, just, I would confirm everything you just said, Aaron. And then to try to put in like slight counterpoint only to try to have a different side of the conversation. Cause I'm pretty much a hundred percent in agreement with you is that if you are going to buy. A service that you're going to [00:24:00] integrate.

Don't assume that like then all your problems are taken right? Like then the integration happens and there is engineering work to be done and maintenance work to be done on the integration. And a lot of these products are amazing. They're unbelievable in the amount of scale that they do globally for the amount of events that they're streaming and tracking the amount of clients that they have.

But every single integration is different and they don't have every answer for every single client. And so there needs to also be at least one technical person to dive in, on the implementation side And help debug it as it happens, because like it's never going to be as smooth as people.

Aaron Glazer: Yeah And it's never the software that brings the success Right. It's the culture around the software that gets developed internally that breeds the success. So you know we're talking a lot about if your culture internally cares a bit the attribution is asking the questions and you're trying to make decisions based off of it. Then you're going to have success in that area you know. If [00:25:00] fundamentally you've got the software there but you're not using it to help make your business better you're not asking questions of it. Then you're really not setting it up for success and it's going to run into a problem at some point that can be a challenge I think of the the bill versus buy When you know if you're building it internally you've got this team they're dedicated to it you know they're gonna they want people using it Whereas if you're external and you're selling software you really have to rely on that company internally to to use it 

Seth Bindernagel: Totally you, you, you have to rely on the company to use it and implement it properly and go all in on the implementation.

Nima Gardideh: Yeah I think that the conventional wisdom here is to buy tools but I think why it's always interesting for especially you know I think in in this specific scenario that Aaron you mentioned it was coming from up top but the scenario where I've also seen is where engineers or even product people get excited about building it in house instead because it's going to be a little bit more fine tuned to the specific business model that they're working on or they just think they're going to have more [00:26:00] granularity or control not having the hindsight that maybe all of us here have had of attempting to build it in house and then becoming if becomes the worst sort of thing ever a year later and I think it is it is a fun and interesting thing because of course no piece of software You both alluded to this where of course no piece of software solves your problems Basically the culture around caring about the problem is what actually makes it work And you can put like two or three business operate together in a way that works you know maybe you connect segment and then Taplytics and blah blah blah and then pull that data into your own databases Do you do your own attribution Who knows but what I've seen a lot and I it was interesting as a big learning for me was actually when I was working with you Aaron is just knowing how hard it is to be able to build these things when we were dealing with all those like scale problems and trying to figure it out uh and then now I go into these companies you know we have tons of clients and they always ask me this question And this [00:27:00] argument is quite often it comes up at a lot of times It's surprising to me that it's an engineering thing where they're like they look at it like of course I can build this You know it's going to take two sprints and we're going to be good And I can't tell you how many times I've had to like show them other companies who've done that gave me the exact same sort of thing and show them Hey look they attempted to build an attribution modeling system themselves on Looker They're still at it It's been six months it's it's an interesting problem but I think it looks easy from the outside when you don't when you haven't dealt with this sort of minute issues that come up and it's way easier when when you've dealt with it before and you know Hey I'm just going to rely on some of these tools to do the basics and then maybe I'll do my own sort of analysis on top 

Aaron Glazer: Well and typically a lot of those like when you're comparing that decision you might be doing it off of kind of the base idea of like oh yeah we could get that going and you probably can get you know one like very small use case of it up and running but to evolve [00:28:00] it to continue growing and to continue like you know keeping up with your competitors that might be using whatever the market leading product is it gets harder and harder and harder because the industry is not staying static in terms of like what do you need to make these decisions and how do you how can you grow How can you put these tactics in market 

Nima Gardideh: Oh yeah Like even an example of experimentation like there's just so many channels Let's say you built one and it was working on web and then now mobile came along or you know TV came along and all this sort of different channels came along And where are you going to invest another like three months of engineering time to support that channel now because you haven't thought about it before yeah It's it's definitely something that I think also because people feel like all these big companies are doing it in house they feel like they should do as well cause if you like go around and you're like Hey I want to know what growth engineering is like. Who or who's really writing about growth engineering. Pinterest a little bit Airbnb a little bit you're not at the scale of these companies to have a whole team dedicated to internal tools. I think that's actually one thing at least for me when I was getting into this was a [00:29:00] blocker Cause it was like I'm reading all this information in the blogosphere and realizing oh maybe I should be totally spinning up my own version of attribution modeling or AB testing or whatever it is and that hindsight was not there where I was like oh yeah the reason they're doing this in house is because they have a thousand engineers I get it now. That's the 

Seth Bindernagel: diffference. 

Yeah, that that is, um, that is definitely the case. I think when you get to these bigger companies, there's also the scale of these huge companies. So like a company, the size of like a Google or a Facebook or whatever, like they are never really going to buy. Oh, they might, but in general, they're not really going to buy their own services because like, they are so big and have so many engineers that like, it's not cost-effective for them to do that as well.

So like, It goes back to where we started this conversation Nima. Like where do you hire? Who at what stage it's sort of similar. Like when do you buy versus when you build every company is going to get big enough that you're going to have to internalize some of these tools because like frankly, like their [00:30:00] systems, uh, the third party systems are not built to scale with like the amount that you might be doing.

One day, you might have data that requirements that you may have one day just may not scale the, the way you have to pay on that data to that third party may not scale. So there's all sorts of issues. With that, that, like, where is that crossover point where like, you are, you have enough people you're big enough, you make enough money.

Your data requirements are so huge that you're just not going to be able to scale, with the service. 

Nima Gardideh: maybe the next one we can tackle And this is something I think has always bothered me especially cause I'm in paid growth. So data is very very prevalent is sort of this like phrase became very popular for awhile and as with all these sort of simplified phrases that tried to explain a concept to people you know like growth hacking for instance which has its own problem has its own problems. I felt that when I at least this is exactly what happened to me I actually was one of those folks that read this phrase and I was like oh I completely understand what this means now and started reading about statistical significance and all this sort of stuff. And it was like well [00:31:00] we don't have significance on this test So we got to wait six more months before we can make a decision on this.

What do you guys think of that phrase And I think the one I like a lot more recently is data informed. Yeah Well how do you think about using data especially in the earlier stages where you don't have you know Facebook level traffic to be testing out millions of hits per minute how do you guys think about the process of making decisions where you have basically incomplete data in front of you. 

Seth Bindernagel: Yeah, that is such a good conversation. Organizations really need to know. Sensitive to making data informed decisions, and not like throw the baby out with the bath water. you could have a great idea. You could do five to 10 customer interviews with like really, really knowledgeable, knowledgeable power users.

You could look at like all the sort of, very subjective, but app store reviews on your app. And you could start to build this body of knowledge of like, what you know is the right thing to do and try. And then you could have a data person that could say like, well, we just don't have the data to prove [00:32:00] that we should do that yet.

So let's, let's just, let's just not do that. And like, you can really stifle innovation if you're always, always trying to turn to data to confirm answers at the same time. Like, I think everybody sort of thinks that they could be like Steve Jobs one day who just was like magical and could like, you know, say what was awesome. And that's obviously not the case. So like, it's this really delicate balance between like knowing your product, just knowing from like a pure, intuitive sense, what the right thing to do is, and then using data to support those things. I mean, we could spend a lot of time talking about this because there's all sorts of stuff that people just don't actually understand how to analyze data.

Even if they're data people, they might be like, oh, that's not stat sig. And I'm like, what do you even really know what stat sig it means? You know? Like it's a very complicated thing to talk about. Sometimes internally and people get very crippled by like, well, the data doesn't say it, so we shouldn't do it. I'm not saying you shouldn't look at data. That's not what I'm saying. I'm just saying like, don't let it stifle innovation.[00:33:00]

Aaron Glazer: Yeah I'll take a different angle at it Um I think a lot of times data-driven means start with data and then lead to the question. Whereas I think the you know my view is that you you'd need to start with the question and then answer it as best as you can and and complete as much of the view as you can using the data knowing that you're never going to have a complete view and you're going to have to make a decision on incomplete data and use the data to answer the question not to start the question So like there's so many there's so much data it's becoming easier and easier to analyze it but without a question then the data is meaningless Like what is it inspiring you to think. How are you like going back to the intellectual curiosity Like what are you curious about with this data. How are you gonna use it to to answer some question I think that is the like the core thing that sometimes can get lost when it's distilled down to just data-driven. 

Seth Bindernagel: You know what I remember, in the early days of working at Mozilla and I'm [00:34:00] talking about like Firefox 1.5 Firefox to Firefox three, where new features developed in the browser, Really creative engineers that just sorta knew the way that the web work and they knew they had to go and build these things.

One of the features that Firefox launched, was private browsing mode, which we all now know as incognito. And w we know it throughout the web because a lot of people use it for whatever reason. And there was no real data to suggest that this was like something that like would have mass adoption, but there were like this incredibly informed group of technologists that understood why, a mode on the browser that would prevent things like cookie tracking would be an advantageous thing for, for certain types of users. and so there's like that interesting combination of like technologists who are really smart and understand their product so well that they can kind of see around a corner and then they can collect data to validate that that feature may or may not have been a good one [00:35:00] versus like, like you said, Aaron, Using data to go in search of a problem, rather than like, thinking about what are you trying to solve.

It just reemphasizes your point that it has to be this balance between like, like there was never the group of like computer scientists and engineers that I work with at most of the lower far and away, the most gifted engineers I've ever worked with. And they craved data as much as anybody, but sometimes it just wasn't there, but they were really gifted at like building a web browser that made the web work.

You know what I mean? So like, it was really interesting to see that happen. And it was at a time when there wasn't data collection, like there is today, you know? So it was really interesting. 

Aaron Glazer: and there's also the other side of it of like you may not have had the direct data of how many people are going to want to use incognito mode but you probably had some of the broader data points of like well how many users do we have. You know what percentage even if it's a qualitative basis what percentage of you do you think might run into this problem or even within the office like you know. What percentage of the people there would want some feature like this and you know you can start putting it together and [00:36:00] and roughly getting sizing that can help inform that question of like should we build it. Cause like the question of should we build it is really like it should be should we allocate capital to this product or this decision. 

Nima Gardideh: I think that's actually a little more about this is I basically have the same belief that I think growth in the beginning might be all about the tactics and getting stuff out But eventually it becomes your job as capital allocation How do you put money behind the right thing So then you can continue growing at a certain rate.

And I think this thing that stood out for me from both of you is this sort of combination of intuition which is like what questions you ask or the data or what you feel like is possible in the technology or whatever And then combining that with data and the way I've been thinking a lot about this as is essentially what I'm what I'm calling sort of cascading conviction which is Hey you have some idea you have a question or you have a product idea and you [00:37:00] got to slowly build conviction that this is a good idea And what you need to do is then try to find available pieces of data to look at first maybe you run a small test to see if if you're directionally correct And as you do those things you're starting to build conviction more and more that this idea is the right idea this sort of approach works very well for our world or my world which is paid marketing.

Aaron Glazer: uh 

Nima Gardideh: Because we never never want to grab all the budget and put it on one idea that we have but it's just not going to work or it could completely screw up the cost dynamics of what we're doing for our clients But there are some ideas that could change everything when it comes to the costs of let's say running ads on Facebook and in fact we'd like discovered at least three of those in the past two years that we've been running the business but the way we've gotten to those it's like

Hey we have the idea We understand the systems extremely well now So we think that if we did this it's going to make it work but there's no way we're going to go to our clients and say we're going to do so we have to slowly say there is an [00:38:00] experiment we're going to run to build conviction that we should do this larger thing that will get us to the goal that we want to get to And I think that approach has made me feel a lot more comfortable when trying to lose things And it also lets our clients as well. 

Seth Bindernagel: It's probably a little bit easier to do in the paid advertising world because it's almost like been, you can point to all these other examples of where you've done this. And you've said like, Hey, like I'll show you where it worked before. So why, you know, why wouldn't you want to do this? And then in the product world, I think you can apply the same principles and it has been applied in, in the product world.

And I'll give you a very specific example. When I was at Twitter, when I first joined there, I was leading marketing for the Twitter platform team. This is a, this can be a very like, open. Role to describe leading marketing for the platform. What does that actually mean? 

well. it means like marketing the Twitter API to developers to build awesome stuff on the API. And it happened to be at a time in 2011. If you remember this timeframe, when we shut down the API to be able to build competing [00:39:00]clients that were building the core user experience that was in direct competition to the Twitter experience. So these are old school days, Twitter, and it was a, it was an all out war that we declared on developers, who we had determined were critically important to us.

And so what we noticed was that aside from building clients, people were using the API to essentially render tweets on webpages. We could just see it. We could just see that like, Hey, people are doing this. So the next step was like, well, why don't we like try to package those API end points into something that's way easier to use to be able to like render tweets on pages.

And we came up with this like toolkit, And then lo and behold, that became the platform from which embedded tweets were shipped. And we shifted that within a year of Epping, like shut down the API to developers and everyone at the time, even internally to Twitter, I was like, ah, I'm betting tweets.

That's not going to be that big. Who wants to embed a tweet and do an article. And you know, now we know how big embedded tweets are, But it all came from us observing these little tiny things [00:40:00] that were happening. And it's very similar to what you're saying, Nima, like even in the paid marketing world, start with these small things.

And if you see the sign of success, build on that, and then, another sign is built on that. That's what we did with, we didn't start by saying, we're going to, we're going to ship embedded tweets to the web. We had to see that people wanted this first. And we had to look for things that were going to help us succeed in the face of us, shutting down a huge part of the API that.

200 different developers to build competing Twitter clients. and so anyway, that that's, that's like, one of these questions where there's someone in the chat room said, what, what is the role of product analytics and finding product market fits? I think if you use analytics and data to do what Aaron said to help you answer a question that you think you need to answer, but understanding what the question is first is how you use analytics to like make product decisions, to make creative decisions and paid marketing, things like that. But like, I, I dunno, it's rare. 

I feel that you like anyone walks in and goes, this is what .We're going to do. [00:41:00] And it's going to be amazing unless you're founding a company. hopefully you guys have some things to respond to. 

Aaron Glazer: But even then like Founding a company like the best founders are the ones that have some some product market fit. Like they're uniquely qualified to build the company that they're building because of some some set of experiences which turns into data basically which which guides them inform and informs them as to why something should be the way it is. 

Nima Gardideh: Yeah it's it's it's an interesting problem is I had a couple other folks on the podcast the other week Uh one of them was a CTO of Webflow And you know I think we are aligned a lot more than I did with Brian who runs Webflow on how to make these steps and decisions because what they had done is basically almost 10 years ago.

Now you're like well this is ridiculous that we don't have a Wiziwig tool to build websites that actually doesn't suck And that has still been the product vision You know it's 10 years It's a very simple thing that they're trying to achieve but it's just happens to be a very hard technical problem.

And it's going to take a [00:42:00] long long time to get it done Right And over the past two years I'm sure you guys both know they're blowing up because they finally hit that inflection point where the product was so good where just no one will use any other tool And they're becoming one of the market leaders And that always screws with my head with these things. 

Conventional wisdom might be that you just kind of ask questions of your customers and you go back and start building things based on that And you slowly build conviction towards the solution you're trying to build but I think Aaron you you just said something that made me think is like maybe their backgrounds or their sort of where they were in the world gave them all that information where by the time they were building that product they're like we know that this is going to be big. 

Aaron Glazer: Yeah I think there's a lot of data that we have that is hard to quantify directly and put down right. Our brains are so good at synthesizing data and creating patterns So they get to to a detriment right. Our brains are programmed in a way to create patterns out of randomized data right And sometimes we can take disparate data points and create a [00:43:00] pattern out of it. And that becomes an insight which is really valuable in building something And sometimes it can actually be a fallacy right Like so sometimes you're creating a pattern out of randomized data where a pattern doesn't exist and that's where that combination. You know what is the question you're trying to ask And then how do you use the data that you have available to you to prove disprove or create sort of proxies or correlated or or causal metrics If you're lucky enough to do that all a causal it gets really hard but how do you create these metrics to guide you along the way. 

Seth Bindernagel: And avoid confirmation bias. [laughing]

Aaron Glazer: Exactly. [laughing] When you're working on things where they could be you know like big bets and you're not gonna know till the end right. Let's say you're working on a big product development You're not going to have until the end And then it takes a long time and you've got to ship it to see if anyone wants to use it. The next step and proved that uh the growth of the idea and continue that investment very much the same as what you're talking about new math of how do you build up the investment build up the case and use the [00:44:00] data and ask the questions of it to prove that. 

Nima Gardideh: Yeah And what's interesting is like I think you can build all the convictions And when you afterwards look at all the data sometimes the answer is no which is I think it's confirmation bias A thing that sets him you mentioned is very important. Like I've seen this in companies that we work with some of them you know think about growth projects. Exactly Like you know we talked about earlier which is everything is almost a capital allocation problem And so they even tried to calculate the potential ROI and all this sort of stuff so that they do net present value on every project essentially which is like a financial commitment Yeah And it's like an interesting thing especially at the there's a couple of companies one of them is about to go public that runs on this model that I respect And before they make these decisions to documents or You know that there's been like weeks of deliberation thoughts put into it And sometimes even they've done all this work they still don't Don't go after the project because after all that's information they still think it's not good enough and they kill [00:45:00] it And I respect those types of moments actually a lot because you have to go against a lot of those like moments in your in your in your intuition You're looking at this this information and then maybe there's a VP behind it that you do trust it to scale your company all the way until this point And you have to make a call uh and then making the right one as in the in the right way might be we're not going to do it even though we've spent four or five weeks trying to build conviction that this is this is the thing to do it to do next And I think it's a very hard thing And I think the.... 

Seth Bindernagel: it's really hard. I mean, to put all your, like, sweat equity into it, I mean, just incredibly hard to like step back and say, like, after doing all this work, having an idea, having enough belief to convince everybody, we should invest in studying that idea, spending all these resources to then study it only to then say like, no, it's not worth it.

I mean, that's a, that's a very difficult request to make it anybody. Right. But like those who can do that, We are adding so much more value than those who then go forward with potentially a [00:46:00] bad decision that can be, months, years of work. The other thing too, which is really interesting and love to hear you guys talking about this is that every decision that is made internal to a company has these downstream consequences that can like have effects for like, Quarters upon quarters.

And it's really important to like, for some of these, like the buyer build discussions, or even like the hiring discussions, like you really need to like, try to make sure you're making the best decision possible because, um, you know, there can be like cycles of having a bad head of product, for example, where you're just like continuing to go down roads and then you have to like, get rid of that person.

And then the next person comes in and they rebuild all the processes. And then like, hopefully that person's correct. Like, there is a lot to be said about trying to like, make sure you make these good decisions and doing everything you possibly can to confirm that these decisions are good and not like moving forward with like what you think you have to do, because it was just an idea. I've seen that a lot at companies where like decisions have these like amazing follow on effects that [00:47:00] really can change the direction of a company. It's sort of a generic comment, but honestly, like if you've ever gone through that pain, it's like, wow, if we just hadn't done that, we would've been okay things like that.

Aaron Glazer: And also you learn so much the indecision and no decision like that phase of like well maybe if we just had a few more data points like it is so powerful to make a decision even if it's the wrong decision versus just sitting there and waiting and 

Seth Bindernagel: Yeah that's totally true. I think one of the things that I always noticed is that, um, when companies start to scale more and more meetings happen on people's calendars and more and more people get invited to meetings and really effective managers of companies are able to figure out ways to like push authority to the edges and empower people so that there isn't this group think group decision-making because of like, if it gets down to the point where like everybody needs to be in a meeting to be able to make a decision and that meeting can't happen until next Thursday.

how many days of productivity are you losing on just being [00:48:00] able to make a decision today. I think if organizations that can effectively scale need to be way better about empowering, like people at the, at the extreme. To be able to make decisions and scale those decisions appropriately.

Like they don't need to be like buying software if they're like a mid-level person, allow them to make decisions. So they don't have to like sit around and like get into this process of, okay, did I, get the 50 people that I, need in this meeting to make a decision? and it always happens. It seems like it always happens. As companies get big, more and more meetings show up on the calendar. Like no one willing to make a decision. 

Nima Gardideh: Yeah I I the way I've always thought about this and I actually believe this is a big thing in growth that is not being discussed as much which is basically organizational management or design I think the only people talked about are probably in Blitzkrieg They talk about it Um and Reid Hoffman's course is a really good one but they're not as tactical They're just kind of talking about the concepts which is what you mentioned is the way I've thought about it as [00:49:00]sort of positive freedoms versus negative freedoms I don't know if you guys are Into philosophy. 

But essentially the idea is that there was a set of freedoms that you get that are sort of bestowed upon you Uh and then there's ones that you are you have to earn I think basically as a founder your job is to kind of get a sense of where you are in the cycle of product market fit. And in the early days you want almost like utilitarian control because you want to make these big changes in the product in order to get the product market fit you cannot have five engineers coming up with a new idea and then building it because you have a lot more context than they do on what's going on with the customers Right But there is a point in time where it's you've discovered the core.

And then now you have to empower everyone to sort of sort of solve all the nooks and crannies of this problem And there's no longer your job to come up with every feature and every big decision it's actually your team's job And this part I think is very hard for founders It's [00:50:00] definitely hard for me and like we had one of those moments today where there was four of us in one call It's like the four busiest people in our company And it took two weeks to schedule this call You know it's just should not happen You know I remember thinking at the end of this call I was like half this team shouldn't have like this should have been done but with two of these folks and and and why do we need to be there to make this decision together even though it's a big one and we should but it felt like we should just empower each other to just do it and trust one another to get there And it's it's such a hard thing to do I think as a founder I think it's hard but uh it is the way 

Aaron Glazer: I think even remote has made it even harder on that side because the cost of a meeting has gone down Right. It's not just you know people have to get together you gotta be in the office the same day It's like literally just a slot on your calendar And the person's gonna figure out how to make it to that meeting So like that that cost of going down makes it so easy to to avoid that in it And as managers leaders whether it be growth or anywhere [00:51:00] else like trying to avoid that cancel that figure out ways to you know get that meeting That's been waiting there for two weeks and just make it disappear is so important

Nima Gardideh: Yeah we started doing a lot of asynchronous decision-making because of that as well where we write our documents basically give the background and then assign the people that we would have put into a meeting And then now we do this thing Well it's a punishment my meeting if you like to not comment on it and help me make a decision that I'm going to punish you by having a meeting about it 

Seth Bindernagel: That's a good way to put it.

Nima Gardideh: But yeah So I think maybe we can talk about next is I think we alluded to it a little bit as finding product market fit but as you probably have all experienced this product market fit is is not a switch At least in my experience I think it's discussed as a switch a lot or oh yeah you have product market fit Great Now it's time to grow but at least in my experience there was different stages of this thing you might have discovered a problem that works but that problem is value for your company's maximum [00:52:00] $50 million in revenue then now you have to go to the next stage of of the problem grow bigger Have you guys think about that I think Seth you're obviously very good at joining these companies that are discovered some some level of product market fit.

Like how do you have you thought about it and I'm sure Aaron you see a lot of this in inside of your data So I'd love to see what you guys think about and especially in the context of growth because you become in charge of essentially expanding product market fit eventually How do you guys think about this problem 

Seth Bindernagel: I think the product market fit, um, I think is probably like your first 10,000 customers. I don't know. I've heard that number used before maybe. And I'm talking about consumer companies, not necessarily a B2B, but I've thought like, it's sort of like, can you, can you S how fast can you scale to 10,000 and then maybe like another, like step function above that, like a hundred thousand where you can start to see that, like, there is core usage and you can start to look at These like kind of standard metrics.

What are the ratios of daily active users that you have that represent the entirety of your monthly active user? And if that's a high percentage, [00:53:00] maybe now, you know, the frequency of usage of your product or, or, or if it's just weekly usage, like what's the ratio of people who are using it every.

Compared to those who use it every month. And you start to look at these things and you're like, okay, cool. We've got something here. And we think we have something that fits now. It's time to scale. And growth is very much like a lagging indicator in this regard, because like, you, you really don't want to pour a bunch of resources into the growth engine, unless you can see some of these core metrics that are Indic indicative of like a very healthy user state model, where you're not churning a bunch of people out of your product.

You're keeping them, you're retaining them. And you're seeing these healthy retention curves. It almost becomes a math problem at that point where you can start to see based on your users, those healthy retention curves beyond eight weeks, beyond 10 weeks, going even further and seeing those monetization curves.

So if you have a, if you have a retention curve, that's flattening out as it heads out to like week eight, week 10 and beyond, you know, you probably have some. The ability to have conviction that you have product market fit. If your, if your curve [00:54:00] is going to zero, you don't have product market fit, right?

Like if your retention curve is not retaining anybody, then you're not, you're not a product that's actually doing super well. So I think that's the thing that we, we just spend a ton of time looking at. And all the companies I've been at is like, how healthy is our retention? How long-term is our retention?

How healthy is our old cohorts? Are they still paying us back? Are we able to move those? Even those oldest cohorts into a healthier state. And if you can start to do all those things with experimentation now, you know, you have, you have the ability to build a growth process. You have the ability to like, do things that can move users into a healthier state inside your user state model.

And if you can't, then you need to do different growth tactics, but you can't. You can't move users who churn. If you're just churning through the internet, then you don't have product market fit. And it's going to be very, very, very difficult to get those users back. I would say [00:55:00] like, don't think about growth until you can see some level of like healthy retention among that core use of early users.

And that that number needs to be big enough, right. It needs to be 10,000 people, a hundred thousand people, something like that. I don't know. Did that answer a question? I know Aaron, what do you think. 

Aaron Glazer: I was going to say I'll still take a look at it on the other dimension of like a point in time So product market fit is a is you know like a single point in time item and you're either going to keep it or you're going to lose and there's this sort of combination of product decisions you're going to have to make And partial part of those decisions are going to be incremental based of like how do we make the product incrementally better.

And then there's going to be those other decisions which are you know like You know I think the market is going here and you're going to have to balance those two as you go along this journey And and as you're thinking about your cohorts and and measuring the impact of your decisions along the way and I think that that's where using the tools like experimentation and others help you make good capital allocation decisions of what you should be building from a [00:56:00] incremental approach or from a you know like net new product approach sort of you know like Twitter platform API for embedded Like that's that's a whole other new area that you have to invest in and believe it but you're going to be watching the metrics and see it grow So so it's really something that like you know you have to balance you have to take on the whole and just rely and be comfortable working in that area of like not having perfect data There's no roadmap for it. 

Seth Bindernagel: What do you guys think is an example of, um, a product that at the time we probably all would have said had product market fit and then it failed. I'm not even sure we can answer that on the spot, I'm thinking about this conversation we're having right now.

And I'm thinking about an announcement that I think apple just made where I believe they're going to like end of life, their speaker, maybe, maybe that was a recent announcement. Their smart speaker, which I think they had hoped would be a competitor to Alexa. Clearly there's like product market fit for smart speakers.

And yet that one didn't go anywhere. And if any company could have taken on Alexa, it could have been apple. I had both of them at one point [00:57:00] and the apple speaker just sucked. It wasn't good. Like Siri was not as good as Alexa, but the fit is there. There's product market fit for a smart speaker to sit on a kitchen, countertop and execution by apple.

Just, I think their end of life thing that I think they, they announced that this happened. I might be wrong. Maybe someone in the comments can confirm that or deny that.

Nima Gardideh: I really believe I would read about it but I completely believe you because I've also used both of these things including a Google home It's

Seth Bindernagel: It's Amazing. And I don't know why Amazon is so much better because you feel like if any two companies should be better, it should be Google and apple, but Amazon, for some reason, Alexa is just better. And I'm not sure why. Anyway, there's probably like scores and scores of products that we could talk about, which are like, yeah, that one could have been amazing. What happened there? 

Nima Gardideh: I think it is a good example there and maybe you know a bunch about Vine. I basically they they have the beginnings of tech talk a long ass time ago and they sold it [00:58:00] too quickly I think And they could have made a big That may have that one of the ones that is still in my head I cannot believe that didn't be as big. 

Seth Bindernagel: I was just thinking about, I was just thinking about Vine the other day. Cause Vine will still pop up. People will still talk about it on social media, like, oh, remember these great Vine videos and they'll like surface them and pass them around. I don't know what else to say about Vine, except that that was just a, colossal screw up that that that product should have scaled and been made a major part of Twitter. and it's a real shame that it didn't.

Nima Gardideh: it's just not going to work even when they really have product market fit in many ways. 

Aaron Glazer: Yeah well oh and even like you know you were going back a few years but there was a you know is a different generation of phone different generation of you know like the usage of the native app on the phone like the there's so many other differences in in like I'm sure I had nothing to do with any of those decisions obviously but I'm sure there was a bunch of data even though it was brought in in terms of like that investment capital [00:59:00] allocation decision of like should we keep putting it in or not And it's interesting looking back at it and saying like oh well this was the early days for it 

Seth Bindernagel: I think one of the um, reasonably involved here cause I was tagged to like lead the marketing for Vine after we brought it inside. so we did a lot of really creative marketing around vine creators who were doing really cool stuff. And I think there is too much observation that the only people using vine were these like hyper creative people that were willing to deal with this like six second looping create like quirkiness and all the like crazy little hacks you had to do just to make like a coolest six second video so that it was almost like the video itself was more impressive because everybody knew how hard it was to make like really interesting videos besides outside of things that just looped.

And I think it was just like more and more conviction was built inside where it was like, this is just a really hard product to use. And w w w I, I think that was definitely a piece of it where people were [01:00:00] kind of like our, our core users don't represent like scale. Our core users represent like. People who are willing to put a ton of effort into making videos, but we want something that like everybody uses.

And like, I think that, I think it just became like a conclusion that like, is everyone just going to go do regular video again? And should we just be focusing on like live video and then like, they move more toward that sort of Periscope decision several years later. But in hindsight they had tikTok they had TikTok in 2013. 

Nima Gardideh: Another thing I want us to touch on before we wrap it up is Uber Because I think that's a it's an interesting problem because if you look at them now Uber eats is bigger which is a totally different business than originally set out set out to build And and I think that is an interesting sort of product market fit discussion because basically their main businesses break even at best or their original business is break even at best in the markets that they're in Maybe they're making a little bit of money on that on the very mature markets where they have a duopoly. [01:01:00]But Uber eats is the one that's really making them making money and it's decently profitable and quite a lot of the markets that they're in.

I always wondered I would love to know how they made that decision Cause I if you if you remember the first team that did it was actually in Toronto which is funny is what Aaron is right now Cause I if you if you remember the first team that did it was actually in Toronto which is funny is what Aaron is right now Uber eats was out of the Toronto team which became you know it's to start launching it everywhere Who is in charge of this type of stuff You know there's a product people as a growth people I'm sure you Seth I've dealt with this a lot And Aaron you you probably work with all these people that are making these decisions is it the founders at the scale in which Uber Eats was launching Uber was already hundreds of people if not thousands of people so I just wonder how how these types of decisions are being made and who who should care about it .

Seth Bindernagel: Yeah, that is a good, that's a good discussion point. And being involved in these discussions at places like Twitter, um, or even early days at Strava. before I left Strava, [01:02:00] there was starting to get to these conversations. But like, I think it's two things. One, there's a product org. That's thinking about how to extend the core use case to like drive, um, repeat usage, to drive engagement loops, to get people like using more things.

Because like sometimes you just run up against your, like the end of your core use case. And then people are kind of like, okay, I've done it all, you know? And like, sit, then you have to start thinking as a product person, like, what's the next thing we could do to extend this use case. I'm not sure if that's how it happened at Uber.

Second to that. There's like a finance team and a corporate strategy team. That's thinking about like, how are they going to diversify their portfolio to bring in more revenue? And those are equally smart people who they may not be product people, but they are like strategy people. And I'm guessing that might be how that decision came to be.

Because like Twitter did this a lot, you know, they'd be like, wow, music is really big. We need to have a music solution. What should we do? And then they go acquire like a small music startup that they hope can take on Spotify and it ends up [01:03:00] not taking on Spotify, but then they do it with like Periscope or one of the orthogonal decisions that they made inside Twitter, which I thought was brilliant was the ad platform team just realized that programmatic buying was going to become big.

And so they went out and they acquired AdMob and it had nothing to do with tweets. AdMob has nothing to do with tweets. I mean, it has something to do in the sense that you can serve ads into a timeline, but like getting users to create content that creates like repeat usage and engagement loops has nothing to do with like programmatic buying of like ad inventory.

And yet that decision, it was that nexus of like, we're going to monetize our product by like streaming ads inside this great content. And then it was like, oh wow. This thing over here is like really taking off and Twitter acquired ad mob, the foreleg programmatic buying even got big and AdMob now represents like a nice, nice little chunk of revenue for Twitter. So I think it was like really gifted, like corporate strategy people as well. That's inside these bigger companies that go [01:04:00] after orthogonally related companies to acquire. 

Aaron Glazer: There's also really great decisions that are made kind of upstream of that that allow them and allow their teams their the safety to try these ideas So in the case of something like Uber eats they had the feature management and experimentation in place to try it in one small area where if it doesn't work out they can kill it without any downside risks.

So everybody could really feel safe about that decision It could push the debt they could probably push authority on that and making the decision decision down try those things And remember Uber tried all those different little experiments I remember like they had a helicopter right And then all of a sudden a helicopter shows up in the app. So like the ability to roll those out at a small level experiment with them and then look at the data and see how does that increase engagement and and you know continue investing more and more and more into it 

Seth Bindernagel: But like Uber's acquisition of, um, what's the, what's the bicycle company they acquired in San Francisco, the red, the red bikes. I can't remember what they are, but they're called. It's not line, but it's limescale. [01:05:00] There an example of where I think they just observed that like mobility in large cities was like migrating toward like individuals riding bikes around because it was way faster than sitting in traffic and way less expensive.

And so, like, I would say that that's probably an example where they didn't experiment into putting bikes on streets. They just went and bought the company that was leading the market and like...

Aaron Glazer: yeah do the math like do that net present value math and then figure out like Hey is this a creative to the company? For sure.

Seth Bindernagel: Correct. Yeah, exactly.

Nima Gardideh: And that was probably my what would have happened That was probably a really good buy. 

Seth Bindernagel: I think that one will come back. 

[MUSIC FADES IN] 

Nima Gardideh: All right. Well thank you so much for doing this with me guys. Thanks everyone for sticking around throughout the whole hour plus hope to see you guys more on this podcast and uh thanks a lot. 

Seth Bindernagel: Okay. Thank you so much. Nima. Thanks everybody.

Nima Gardideh: Thanks for listening to our show. Get our episodes. As soon as they're released, just tap that follow or subscribe button, wherever you get your [01:06:00] podcasts.

Plus if you want to join our live discussions where you can ask us questions, as we record, sign up at pearmill.com/hypergrowth-podcast. We'll see you on the next episode, on The Hypergrowth Experience.

[MUSIC FADES OUT]