S5 | Episode 9 | Julien Lavigne du Cadet of Zoe: From 0 to 1 Million Users in 36 Hours: Building A Platform That Can Rapidly Evolve

He also dives into growing the culture within Zoe because they've scaled massively and we dive into something I think is really exciting, but they built a mobile product within five days and it had a million users within 36 hours with that. I think that's one of the more exciting Businesses and products that we've had on the show today.

So hopefully you enjoy it Welcome Julian. Thanks for joining us. Thank you. It's great to be here. Do you want to start off by telling us a bit about yourself? Zoe and your journey within the business as well. Yeah, of course So i'm leading the engineering team at zoe i've been a company for about three years Zoe is a personalized nutrition startup.

We did some medical studies many many years ago and And our hypothesis at the time was that we all respond to food very, very differently. And that's basically kind of what we established. So if you and I are eating exactly the same meal, actually our biomarkers are going to vary a lot. So we're all unique.

So I can hit, you know, muffin and have like a very good response in terms of glucose. You have a terrible one. Or the opposite, most likely the opposite actually. And yeah, we've built a product basically to help people understand how to respond to food and give them personalized advice. So that's what we do as a company.

My background before Zoe was a lot in very different industry, quantitative trading, mostly at tech. Spend a lot of time building distributed systems or very low latency trading system, as in you're scrapping microseconds out of things or working with a huge data set, you know, multi petabyte you know, running kind of analysis and, and building platforms supporting that very different experience from what I'm doing today at Zoe.

Yeah. So one of the things where we start putting this together was that you mentioned when you joined Zoe, there was no products and only a vision. How do you kind of, can you talk us through how you worked purely off just vision basically and made? Yes, so I I remember that, you know many years ago The founders of Zoe, this is someone that was someone that I knew from from before Reach out with this this grand idea and that was kind of cool Five years ago.

And indeed, the vision kind of stayed the same since then. So since the beginning they had this idea that, you know, we all respond to food differently. We're going to build a product around around that. So when I joined almost two years later, which was kind of three years ago, we were still at a point where indeed there were no product.

We had done the medical studies and now that was not just purely vision. That was something we had established with collecting, you know, very, very serious data. We were in the process of writing scientific papers And and the challenge really was yeah, how do you take that and build a product out of it?

So Our view at time was well to pro to recreate kind of the same experience that the people had with the medical studies, we will need to ship them some kind of kit so that they get tested. But it also means it's very complicated to build the first version of your product because how do you test that when you provide results to people, they're going to like it.

So we actually kind of decided to go with a very Very with a free version of a product where you could just kind of, you know, enter a few data about yourself. And this is really how we got to kind of test one of the core features that we build later, which is the idea that we can score any any meal for you.

So, you know, You log a meal and we're going to say, Hey, this is that good and not for you. So we tested that with with a free version of the app. But then quickly, we always had in mind that, you know, we really want that to be personalized to you. And if it's going to be personalized to you, we really need to test your, your biology.

So that vision was kind of very clear in our, in our head and, you know, our journey has really be like, how can we try some of those things without building everything? So a lot of startup talks about MVP and for us to kind of first version of the MVP is kind of provide the kind of what we call the results experience, what you're going to receive after testing actually without without building the test.

But it was hard because. You know, it's very different experience than when you have actually the full test. So at many point in time, we had you know, kind of leap of face. We, we knew where we wanted to go. We had this kind of vision and that was pretty much, okay, let's let's test what we can.

But then also, let's not forget that this is exactly where we want to go and where we want and what we want to achieve. So we've kind of iterated in, in several ways. And I think a bit after the start of COVID, I think it was in August the same year, we actually started you know, we launched our product.

I actually kind of put it out there. selling it to people with the kids. But initially, you know, the first few weeks were selling like 20 kits or something like that, literally because we had to assemble a part of the result experience, like, you know, the PDF or sending to people like manually, like we had some people that are just kind of a drag and dropping you know, text PDF and sending them to people like the app was also kind of very basics.

Yeah. So again, kind of iterating, trying to improve on that, getting some customer feedback but always with this very clear no star that our testing experience was extremely important. Yeah, perfect. And I think one of the things as well, so you've, the product's grown, the business has grown, obviously, how's the culture kind of been built alongside that?

So I feel like we are a very lucky company in many ways because or our funders are like incredible people and they are incredibly supportive individual and they've hired people in very, very similar ways. So they've, they've had, you know, previous company that, you know, succeeded And they always cared a lot about culture.

So they always kind of hired people that also cared a lot about culture. And I would definitely kind of consider myself being part of that group. This is, you know, I love technology, but I love also kind of building teams where people can really do great work together. But culture is really a journey.

And you know, the culture has evolved a lot indeed between when we were 30 people, three years ago in an office we're definitely not remote first to what we are today, which is about 170 distributed across Europe and the U S. So that's been like a very interesting journey. And I certainly remember, for example, when, when COVID did like everybody else we decided to.

There was kind of no other options, but maybe contrary to a few you know, a lot of other companies. I would say about kind of six to nine months into the pandemic, we actually make a very deliberate choice to go remote first and stay remote first. So this kind of desire to actually have a very clear culture, like make choices, say, Hey, okay, we're not necessarily for kind of everyone because we're going to kind of optimize more for for something.

So I think we've always kind of pushed on that. We've always also tried to clarify what we're. Or values or operating system. What are the things that we care about? So if you know if you were join us at Zoe when you're on board, like we've got like documents to kind of explain, to expose that to you.

And then we try to build, you know, really good feedback loops also to kind of reinforce the traits we, we care about. So, you know, that's part of our performance review or 360, you know, we ask, how are you doing on those different values, constantly So, but that has been something that has been like evolving and will keep evolving, you know, as as, as the company grow, but we've always kind of invested in that because we always wanted to have a company where we all kind of enjoyed working, not just be spectators and just, you know, let it happen.

Yeah. Perfect. Great. I think one of the things as well, Chris, I guess you'd call yourself honestly, though, you're VP of engineering, you're a software engineer by trade and we'll always want to sort of understand how do you as a software engineer, how do you. Become the best at what you do. How do you kind of become more efficient and yeah, better in your purpose.

So. Software engineering is a journey. And I, I do remember when I when I started many, many years ago being kind of a very rugged about it in the sense, you know, I arrived like, oh, I know how to do all that stuff. I'm probably very great at it. And then a few years later you realize actually how much you don't know and how much you need to learn.

And. I think that's not going to be new for kind of anyone can listening to this podcast, but software engineering, I think you really need to like learning. If, if you don't, then, you know, everybody else kind of, you know, gets better and kind of you, you stagnate. So I do believe that you require a certain degree of passion to be to be great in, in that field.

So that's something that, you know, very early in my career. I was consuming all, you know, the new site that you could imagine about software engineering and some of those things, like I had no idea. how they would become useful to me. It was pretty much, Oh, let's, let's get interested in all those things.

Let's build some kind of mental model about kind of different technology, how to interact with each other, front end, back end, database et cetera, et cetera. And I found that to be incredibly useful early on in my career, because You know, at the beginning, I was reading some things I just did not understand, but then it start clicking and you start finding some, some patterns.

And that's definitely something that I would take Recommend people doing just, just read a lot. Obviously reading is great, but you also need to exercise your craft. So try new things. And I think especially, you know, early in your career like I'm a big proponent of software craftsmanship. So how do you write that code better?

And I found over the years that many people are maybe a bit passive, you know, to join a team and the team has some kind of habits on how they do things. Yeah. And they don't necessarily challenge that. But I would say the opposite. Go and challenge that. Try different things. Try different ways of, you know, writing code for architecturing systems.

See what happens. See what you learn. It's by learning kind of different things that you will actually start to understand, you know, when the tools make sense and when actually you should probably not do it. And, you know, it's, it's, it's kind of all about it's all about trade off. And maybe the last thing that comes to mind, which is an advice I do give to junior engineers in particular at Zoe is that at some point and the sooner the better you need to go really in depth into something and it doesn't really matter what that is.

Like you can go very in depth in mobile development. You can go very in depth in backend development, et cetera. It doesn't really matter, but if you do it at least once, you'll realize that you develop the kind of learning skills that you can now reapply to any topic. And that really, really boosts your confidence.

And that means you're not afraid of jumping in something entirely new because you knew, you know, that you, you know, in the past you were able to learn about something very complicated. So that's just another very complicated things that that you need to learn and just a repetition makes it go faster and faster basically.

Yeah. I must admit, I very much. Believe in growth mindset and learning, but I am not jealous of what a software engineer has to do. There'll be a new JavaScript framework. There'll be a new function on AWS. It's just constantly changing. It would blow, it would blow my mind and it would tire me. I think so fair play to you for having that continuous mentality.

I think one thing to really dive into as well, cause your product is always really unique. I'd love to know how you've kind of got in front of customers to sort of show them the product exists, but maybe they're not. searching for it. They're not necessarily going out there and putting in that into Google or whatever.

Indeed. So maybe like if I summarize first a bit better what our product is about. So fundamentally we sell you this kit and this kit has different parts of it. So you have specific muffins that have like very you know, a known a micronutrients amount. So you're going to have, eat those muffins at certain point in time while you're wearing a cgm continuous glucose monitor after eating the muffins you're going to take a blood sample with a finger prick and then later you're going to take a stool sample and basically we can with all of that we can kind of baseline you and tell tell you how you're going to respond to food and after that really kind of give you all the personal advice about how to eat better how to be healthier so indeed back to your point Not that many people are looking for, you know, personalized nutrition program.

At least, you know, when we tried to do some adverts on Google, like, we pretty quickly found out that no one was clicking. So for us, there's been a different part. In general, we would say that there is a trends towards kind of everything that is gut health related. So we're definitely kind of using that to kind of support or growth, but we've also done kind of a lot of work, especially we've got like wonderful people in, in branding, working with the press to try to just get our ideas out, like not even kind of selling Zoe directly, but just.

You know, talked about those issues, explained to people that actually were kind of all unique. So we've had had incredible press coverage in in the last few years, and some of our scientists are really kind of your key influencers. You know, they, they you know, different people love having them kind of there.

writing article about it. Our founder Tim Spector has written three books on the topic. The latest one was released about a a month ago and something a lot of people actually just read the book because they're really interested in about that area. I know through all those mechanism, basically we get we get Zoe out there but you know, even with that, it's indeed, it's, it can quite complicated to for us to acquire customers because.

Our kit is not cheap. It's currently like two 60 pounds to buy it plus a subscription. So you still have quite a lot of work to do from, Hey, this sounds like a nice idea to actually do. I want to buy to buy the thing. And we've kind of developed some kind of a marketing journey where people go through some kind of quiz where we try to explain to them like the value of the product by having like very kind of concrete examples of, you know, what what does this mean to be like a high blood sugar responder?

What is the impact? We expose them to our science. I think that's something that is really important as a company. We really have about 20 papers have been published in some of the major you know, major scientific press like, you know, the Lancet or Nature Medicine. Like, it's really, really hard to get into those place.

And, you know, quite early in the process, we kind of tell them, hey, this is, you know, we're actually giving back to the community. This is what we found, this is peer reviewed so, you know, it's not just, you know, don't just believe us, actually, you can believe the people that reviewed those papers, that the science is, is solid.

And yeah, basically there is this investment to really try to kind of, you know, get people to understand that we're just not a small product like these that can buy and forget about. But it's almost like a new philosophy where they really need to get into kind of into what we offering, which is basically very different from most of the diets.

Most of the diets are very restrictive. We're pretty much the opposite. We really want to provide you with more options. So yeah, we, you know, we try to explore that in various mechanisms and we have got our podcast which has actually really, really successful in the UK and the U S where we invite scientists and talk about a lot of things, which has intermittent fasting or.

You know vitamins are the good or not for you and like many of the topics. So if you're interested in that area, just listen to podcasts at least one episode to to learn about some of that stuff. Cool. Perfect. Yeah. I think one of the things that actually first came across Zoe was your the COVID app.

And I believe if i'm correct that you basically went from zero users to a million users in like 36 hours with that app. You How on earth did you cope? And then also, how did you manage that? Because that's like, huge scale, right? In the blink of an eye. Yes so Indeed, two and a half years ago We took a bit of a detour At that point in time We didn't have a product.

Our nutrition product was not launched. So it was what was it, March 2020 when when we all went to into lockdown. And you know, we were working with the scientists already. Our funder, Tim Spector is running something called the twin registry, which is A registry of about 10, 000 twins are the most studied people on earth whatever you want to understand the impact of genetics on, on something you basically run a study on those people.

And you know, we had this idea that, oh, maybe we could help with the pandemic, but initially our plan was, well, let's build an app just for the twin registry so that they can log their symptom. And based on that, we'll see, maybe we'll find something useful. And we really run that as a hackathon.

So the interesting thing is that we went from zero to 1 million users, 36 hours, but actually. The project had started five days before so if I remember correctly, we're kind of thursday morning. We're like, hey, let's do a hackathon Can we build a mobile app in two days? On the friday like midnight we had like a first version of a mobile app Like we were happy to kind of push it on the app store and play store And then realize that oh It's an app about COVID.

All the teams are actually validating the new apps. All went remote. Like it's a bit, everything has stopped down at Apple and Google. And on top of that, since it's an app about COVID, you're basically raising all the red flag because, you know, spam and and kind of everything you can imagine. We still managed to get it out.

And I think we launched it and did on the on Tuesday morning and 36 hours, one, 1 million users were kind of incredibly lucky in the sense also that was kind of the the right timing. But yeah, the whole thing has been like, you know, it's, it's been compressed. Like how do you launch a product in five days and have like this, this scale in, interestingly, the scale element of it, Never was like the biggest problem.

Because, so, you know, it was like a mesmerizing on the day of the launch to see kind of the numbers. But those days versus kind of, you know, 15 years ago, you can actually rely on cloud technology to do a fair amount of scaling for you. So I remember on, on the second day. So I think, you know, at the end of the first day, we had like, you know, 200, 000 people.

I was like basically 10 hours after launch. The next day in the morning, everything crashed like 8 a. m. I was basically were somewhere on BBC or something like everybody was kind of installing the app. But again niceties of cloud technology, which just went into our cloud provider. You know, the problem we had at that point was just the database was just did not have enough resource.

We literally took the biggest one that could provide. So we went from whatever we had at the time, eight CPU, 16 gigabyte of memory to Yeah, give us the one with 96 CPUs just in case. And then the front end was actually scaling automatically. So so that part actually in some ways was kind of way easier than than many people anticipate.

Also, because even though the numbers of users are quite high, one million users, this wasn't that that you were only using like once a day to report your symptoms. It's not like continuously scrolling, looking at stuff. So that made managing the scale easier, but still we had some very interesting problems such as the dreaded, TV spike which is the next kind of several weeks.

When our funder again, team was every now and then going on TV and that whenever that happens you know, you go from whatever 200 queries per second to 2000 queries per second, literally in the span of less than 10 seconds. And your great and wonderful cloud autoscaling technologies just doesn't react fast enough.

You know, it takes one minute, two minutes to get more capacity. So you get this big latency spike or everybody isn't happy. And that is a problem. It took us, you know, a fair amount of time to kind of optimize and then kind of several weeks to get right. But yeah, so technology wise, actually it was not the biggest challenge.

The biggest challenge was actually the kind of human element because there was no plan. When we launched that, there was pretty much, you know, we think this is going to be useful. And then with that scale overnight, we've pivoted the whole company, which at the time was about 30 people towards this.

Some of the things that broke is like overnight. We had like, I think on the, on day three, We had like a thousand email of people, you know asking stuff and that just continued for the next several day. So we ended up Three four people that were doing like totally different jobs at Zoe in theory Just going through emails and you know doing stuff like as I said a whole company pivoted towards that on the engineering team like for about two weeks, we didn't sleep a lot because You The app was really, really scrappy.

Like you had no idea as in we launched without the ability to reset a password at the time. So, you know, a lot of those emails actually, Oh, I've lost my password. Like that was the kind of scrappiness we talked about. So there was a lot of work to not even get to a point where it was kind of great.

Just get to a point where it was just, you know, barely okay. But yeah, eventually eventually this kind of normalized and you know, they still exist. This is not a health study. We've launched something called a intermittent fasting study. Like a few, a few weeks ago 150, 000 people joined the study to help us understand better intermittent fasting.

So that was kind of never, The plan in Zoe but now that's an important part of of what we do. And there's actually a lot of synergy in with what we do on the nutrition side. So it's been an interesting journey. Fair play for building an app and getting a million users within a week. Like that's crazy.

I don't think there's anyone else on this podcast that can say that. And I couldn't name another business that's done that. Well, I'll tell you the funny thing, which is So initially in this hackathon, there was actually six of us, four software engineers, one product designer, and our CEO actually, that was looking at the scientific part, like what question we want to ask, and also the legal part.

And I was actually one of the software engineers, and I had not done any serious coding for some amount of time. That was actually my first mobile app ever. So that's a funny thing, which is my first mobile app, you know, just reached 1 million users. I will never, ever touch any other mobile app, because I can only do worse now.

You can't beat it, there's no chance. Brilliant. Obviously you touched upon some stuff around scaling with that, and you made it sound relatively easy, but I'm sure there's Some challenges that you've kind of come through when it comes to the complexities of scaling as well So you want to talk through any what was kind of the biggest challenge?

when it comes to to grow in the product or scale in this business, I Mean the the biggest challenge I think in a in a startup well, I guess part of it just depends what what you're building, but To a large extent is really the first step is always kind of Trying to find product market fit is how do you build a product that people really, really care about?

And it's interesting because on one side, we had what we did on COVID where it was very clear from day one that there were product market fit, even though we had done zero research, we just launched the thing, had no idea what we were doing, but the market was pulling you. And then on the other hand, you know, on the nutrition side, kind of very different stories, It's an expensive product that we're selling.

So you have to provide a lot of value. You have to really kind of try to understand what what matters to people. And and that's really kind of, you know, one of the biggest challenge as a startup is what are the right trade off around that? You know, Do you need to be scrappy? On the contrary, do you need to provide something kind of really like great experience because it's very expensive?

What shortcut do you take on the technology side? This is one of the thing I've kind of learn maybe the hard way by joining ZOE. When I joined, I was probably a bit naive saying I joined ZOE because you know, I wanted to join a company early because I wanted to really be able to influence directions of many things.

You know, I had all those mistakes I had done previously in different company, all those things. I was like, no, we'll totally never do that again. But then also what I probably did not fully appreciate at the time is that, you know, in all those companies you know, we're not really fighting for survival and then you join a startup, you're actually fighting for survival.

And now suddenly all your great principles, all the mistakes that you don't want to do. Do again, you're like, yeah, actually we do have to take all those shortcuts because, hey, how do we get to the next stage? How do we earn the right to you know, build like a this, this community have people that kind of buy our product earned the right to kind of raise money with with VCs.

And that's kind of the really hard part. And I think it was hard for me, but sometimes it's also hard for engineers because you, you know, you know that sometimes what you're doing Is not necessarily kind of, you know, the best you could do. But sometimes, you know, it's kind of the Pareto law, right?

You get 80 percent of the value from, from 20 percent of the things you do. And it's more questions of, you know, trying to really zoom in on those 20 percent and not focusing too much on, on, on the rest. So I think that's, that's a very interesting kind of a challenge in terms of scaling a startup. And also you've got, You know, you always have all those fires around you, and it's really like you're zooming in on the one that are kind of life threatening, and the rest, just let them happen.

So, you have to be comfortable with that. You have to obviously pick your battle, and And really try to focus on the right thing. And at the same time, you need to invest enough in certain area that, you know, you're not going to die a year from now because, you know, you've got too much debt decked.

Previously I, I worked for a company called Criteo. That was like a big success story in France. And one of our founder Jonathan was also from, from this company. Before I joined they had reached a point where it had so much technical debt. That for almost a year, they could not release anything.

And actually we're very lucky at Zoe because Jonathan was one of our funders has been kind of burned by that. So is it really appreciates that, you know, technical depth is a danger. And, you know, it's always hard to find right balance, but, you know, we try to make sure that what we do today will still survive, you know, the test of time without like totally going overboard on, you know, investing where we should not go.

Yeah, great. I think one thing to dive into then is just that you've obviously built a system that's had to evolve rapidly. But it's obviously got to be, the system itself has to be, you know, well, relatively well maintained as you're doing that. So, how did you work through that? How do you build something that has to, you know, be able to adopt and evolve quickly?

Yeah, so I guess for the kind of software we, we, we build which I guess for, for Zoo, we, we've got you know, Significant backend and then we've got mobile apps. I think on the backend side, like I'm generally more like in the favors of kind of building microservices. Not necessarily kind of on day one, but you know, as you start scaling your engineering team, you really want to go towards some kind of architecture where you can actually, you know, take a small part, replace it if necessary without having kind of to break everything and, and, and redo everything.

And I have found that if you can really kind of decoupled your system, that's kind of that's kind of easier. That's also easier because that generally means that the cognitive load for engineers is way smaller. And so, you know, when, when you join a company you know, you just need to know about that part to be able to kind of, you know, improve improve it and kind of build features.

Almost every company I've joined the last kind of 15 years, we'll always start in with some. You know, very ugly monolith and kind of evolve more towards that And and that's really kind of the main reason which is when you when you join, that monolith, you know, nothing about it Like, you know, you you need to be productive in a few weeks time It's just really really hard.

It's just really really hard because you never know what you're gonna break because you don't understand the boundaries people have generally speaking like Great intentions. So, you know the people that kind of start building that they're always like they're always like, oh, you know You will have that thing.

That's very close separation monoliths and that other thing but then as companies scale You lose that and then everything starts to become this big ball of mud and new engineers So Zoe was not different. And you know, when I joined, we had a monolith. We slowly started carrying a project. Chipping away at it, kind of creating microservices you know, never in the, never with the idea of kind of rewriting everything, in my opinion, that, that, that doesn't work, but more like, Hey, we're going to start building this kind of new features instead of putting it here, let's create something that is a bit kind of isolated.

So we've kind of, you know, always kind of approach it from that angle. And I guess the other thing is that we've always tried to really, short term delay time. So what can you do to get your thing out as quickly as possible? Which is both true on the back end, but has also been true for us on the front end.

So on the front end, for example, at the beginning, We had full native apps. So one was built with Swift and the other one was built with Kotlin. But that also meant two code bays. Which meant, you know, we're not very rich in terms of how many engineers we could hire at the time. So we had one mobile engineers for iOS, one mobile engineers for Android.

And this is a very uncomfortable situation to be in because, you know, if one of those engineers leave or just go on holiday or is sick, it's Everything stops. So we made some kind of hard choice quite early on when, when we joined, when, when I joined one was Hey, actually let's just stop entirely building anything on, on Android.

Why? Because we were still at the point where the goal was learning. We didn't need to have like the two platforms to learn. We could just do it on iOS, which for us was a bigger market. So we're mainly targeting the US. So we don't get a 60 percent market for, for the kind of thing we're doing. You know, why pay twice the amount?

To not learn more because, you know, we could learn everything we need in an iOS. So, you know, first thing we did, and then when eventually we decided that, oh, we actually want to have an Android app because, you know, we do want to serve that market, that's where we decided to Move away from full native and go towards react native to have cross platform and you know, it's hard to know if that was the the right timing but you know, we got some really nice things from from that which is one code base So that means, you know, you generally go faster at kind of equivalent amount of resources And also very interesting thing we gained the ability to do over the year update, which is not something you can do with native apps.

So, you know, we can literally do a release any time of the day. Whereas normal flow is that you have to get approval from Apple or Android. And that's also kind of, you know, very comfortable in terms of lead time, because that means suddenly you feel a bit more relaxed. You know, if you introduce a bug, You can fix it very, very quickly.

Whereas if you introduce a critical bug in a native app, you know, you can have at best a 36 hour cycle most of the time to get something else out, which suddenly means you need to invest way more in QA than we would in react native because, you know, we have this ability to fix things faster. So there's a bit of a shift in mindset.

Doesn't mean, you know, we want to ship things are broken, but, you know, we definitely a bit more relaxed than when we're doing native development. So I think for us kind of focusing on lead time focusing on our ability to recover quickly when there are problems has been important. I know those are generally falls under what's called the Dora metrics, which is definitely something we're looking at regularly.

Obviously with your career, it sounds like you've had some great successes, but what has been one of the biggest challenges in your career to date? There are so many challenges. I think it's, it's interesting because as I reflect a bit on that you know, you can look backward and you look a lot of interesting technical challenges, but I think most of the people that evolves toward kind of leadership and leading people.

will tell you that actually, you know, building software is reasonably easy versus versus leading people. And you know, I feel for the kind of software we build, like, you know, I feel pretty confident that we can build it on the, on the technical side, but yeah, people is always hard. Change management is, is the biggest thing is like, how do you get a group of people from A to B?

How do you really kind of motivate people? How do you kind of, you know, reconcile very, very different point of view. And you know, you have your own point of view. It doesn't mean they're right. So, you know, for example, we've got an interesting discussion, the moment that Zoe, which is what is the right amount of kind of upfront investment required in.

product discovery. So how much do you kind of test ideas without writing the first line of code? And there is very interesting tension here that is kind of hard to resolve because Obviously writing code is quite expensive. So you do want to avoid doing that if if you have other ways to kind of test your ideas, but then on the other hand, I would be more on the on the camp of that will only tell you so much because.

You know, we're all very bad at predicting our own behavior. So at some point, you actually need to have the thing in your hand and test it. And, you know, where, where do you put the cursor? Those are actually the biggest challenges. Like, how do you get a group of people to agree on kind of your principles that that you want to implement going forward?

How do you kind of create that alignment which is kind of essential towards kind of, you know, going fast and people knowing what what they should optimize for. And you know, this is kind of an internal repetition, like you, you do it on one topic, but then there's another topic with like a different set of people.

And that just requires a lot of effort and, you know, sometimes succeed. Sometimes you don't again, you tend to be always very naive. At least that was my case. When you approach this problem initially, like, Oh, that must certainly be a super easy to align all those people. But no, whenever you men are involved, actually, this is not that easy.

So I think that's, that's one of the biggest challenge and you know, something that I'm still faced with, like, you know, And on, you know, everything that is on the technology side, back to what I was saying, is actually generally way easier than that. The final question from us then is what piece of advice would you give someone who's looking to join a startup?

So, I think, Joining a startup is kind of adopting adopting a certain kind of style of working. And I actually do think that joining, like, you know, a startup that is kind of going fast, et cetera, like the, the, the years of experience can double. But that's also Not for everybody and also kind of depends on what the startup is How much support you can you can get?

so If you are, you know, super super junior and you join the startup where there is only super super junior people that's going to be hard because You will be exposed to a lot of super interesting problem But also they won't be that many people to help you kind of grow and look at you know, what, what great look like on, on the contrary, there's kind of a few senior people that can mentor you.

That's that's a very, very interesting experience. So I think, you know, joining a startup, my advice is that there's actually, you know, startup and startup, and it's, it's more like a matter of kind of really finding one startup that is, you know, Going to match where you are in your career What are you kind of aspirations, but it's it's a great experience.

It's a great experience in particular if You really want to kind of grow in totally different areas And if you care like purely about how I want to be like the best software engineer ever and you No, don't care about kind of, how do you build like a great product? Or, you know, how do you build team?

Maybe you will be you know, better suited to join kind of a more mature company where you will be able to see what, you know, great processes look like. But if you like kind of, you know, growing and be stretching totally kind of different direction, if you like to take a problem that is like fundamentally quite hard to solve, where there's a lot of ambiguity and just kind of feel the gaps and get a lot of things done.

I guess the startup is probably a place that you will enjoy more. So it's kind of quite related to the degree of ownership that you like taking. If you really like taking a lot of ownership, make things happen. I think startup are a great place. If you know, you're more like, Oh, you know, I want to focus on this super complicated problem.

I don't want to worry about anything else. Then, you know, maybe there are some some other place that that would suit you better. So I think, you know, my, my advice is, is be creative. Basically that try to be a bit you know, conscious about what you want from the experience, what you're looking for.

And then, you know, there's a lot of startups to find one that can offer you know, good supporting environment versus what you're aiming to do. That's all wrap it up there, but thanks for joining us, Julien, really appreciate it. Thank you very much.