How To Start A Dev Tools Company with Nicolas Dessaigne | Startup School
32:56 Share

Hello, my name is Nicole do sania and am a group partner. Here I see before that, I was the founder and C E, O of a golia. Allia is a search A P I used by millions of depth to build a great search experience in their apps and websites.

And it's live on today because today we are going to speak about death tools and how to start a death tools company. So here's what's working to cover. First, we discuss about the funding team and how to find your idea.

Then how to start the company from the prototype to the M V P, and finally will dive into the go to market, including cells and marketing bd def tools. But first, what is a def tool? So a def tool is software that is used by developers to help them be the and that includes a lot of aspects from coding, testing, diverging, documenting, deploying, running.

It's all of the canada aspects of that job into broad category. Then you can think about I D is like V S card, A P, I like trip to you or algol. A my company library and frameworks like react no G S or log chain, even infrastructure or clothes services like s over so and tools doctor, you get IT is very diverse and that I see we've supported many of these other tools, hundreds of them, and already have two public companies.

Github, the open source tab and page duty that started as a simple alerting system and now is used by half to fourteen, five hundred companies and many users you never know about stripe docker hahaha u superb signal potok apple airport and also are going, of course, and we've learned a lot from them. So let's dating and start at the funding stage. Well, first need the funding team and the key uh aspect of their tool l is that you are going to build a technical product for developers.

Obviously, you need to be a developer yourself. Most of a worker funding team for live tools are all developer teams. And the great thing about death tool is that developers actually using death tools everyday.

So when you are building your death tools, you are basically helping yourself in your action of building a product. So let's take a step back and chat quickly about what are good and bad, the two ideas. And as a profess, let me start by telling you that with the advent of ms.

And A, I gentle, it's actually very difficult. Know for sure today what's a good or bad idea? The bad ideas, the topic ideas of yesterday could actually be possible to make successful.

Now, thanks to elms, some ideas, actually more difficult, everyday, are going to work on things they are actually not going to like, like documentation, Q A, testing. And these become abuse ideas for them to fix for us. The problem here is that they are not the only one to think about IT.

And that creates a lot of notes, and that's why there are so many of these on the market. Do you really want to build the next and the next tool for auth? I don't know. Another way to look at IT is the difference between the time and your own time ideas.

So the q testing, all of these ideas are build time ideas, things, tools you are going to use when you are reading the software are IT happens that these ideas are mostly nice to have. You could build your product with them. What's more interesting are one time ideas.

why? Because at one time, IT becomes really critical. It's most half products. If you bet on A P, I, for example, you cannot run your own product. If it's done, that's why it's more critical and it's very Better idea support of tools.

The user advantage of this home time ideas is usage because the more your customers are going to go, usually, the more they are going to usual product. So your incentives, I think about scribe, for example, if that customers sell more product, they make more money, everyone is happy. Much Better idea.

And there are a quick category ah that began touch phone is libraries and frameworks these can be great at is usually open and source, but they can also be very changing to monitor ze. Think about pandas as, for example, very differ. To make money out of that.

The one way you can monitise ideas is by offering your hosting service. Think about next G, S. And versa, for example.

OK. Last point I wanted to touch on this idea, which is the L M. And A I trend. Actually, with this change in the world, there are many new death tools being created.

And it's also it's so exciting to walk on the death tool that's going to help these companies meeting in ms, but it's also the time when the market is very nicer, so annoyed. Great idea is a lam observably. Well, uh, it's so just right.

Anyone building with at them, they want to know how they are doing. They want to be able to evaluate them. Guess what? Many companies thinks the same thing. And that's why you have dozens, if not hundreds, of such companies out there. IT doesn an that is a bad idea.

But I think that if you are starting in the space, you need to have a clear idea of how you are going to define chat yourself or are you going to win against all of these pets? okay? Which leads me to a few mistakes.

People often do. The first one is waiting for the perfect idea. You are going to wait for a very long time. Let me go back to the observed ideas.

I would actually inquire you, is that to do idea are just that? That's okay, even if you don't cheat yourself away to IT, that's you learn as you go, and eventually you be able to read up or you will change your idea, which is me to my second mistake, sticking with the wrong idea for too long you may not realize, but fifty percent of the companies in Y. C.

Eventually be vote from their first idea. It's huge. And it's also too for death tools. Finally, last mistake, thinking, you need a business founder in the team.

I know that what I thought when I need a, we need a to take of orders, but across our mind, we don't know how to sell. We need to hire all to find some mocco founder who knows about sale. No, it's actually with your than you think to to learn how to sell your product.

Actually, we check the numbers, seventy four percent of ways the two companies add only that go founders, and that's compared to forty five percent for all of the use of companies. okay. So now that you have a team and an idea, where do you start whether there is only two things you have to do, build and talk with users.

There is no specific order. You can just stop by talking with users because are going to help you figure out if your interview is good. You can stop by building a little time so you can get your feed back.

Let's speak about that prototype. The most important thing here is not to over engineer IT. You should do that quick and dirty. And I know that I can be very difficult to for some developers out there is for experience engineers who pride themselves on the of their, but the tools that the most important thing at that stage is to eat rate as fast as you can. For some, I would assume that you hall away ninety percent of all the code.

You right? So your goal at that state is just to identify the ten percent of what your building that's actually valuable as fast as possible. And then later on, you can eat or these ten percent only and avoid wasting a lot of time as we ate.

Of course, you want to speak with users and get feedback. It's completely okay to show them a prototype. Please don't wait to have a perfect product, just shows them a prototype and stop gearing feedback as earlier as you can. And that will eventually lead you to A M V P, A minimal viable product. And this were the very important viable you need to provide the value to your customers.

It's much Better to be ten Better on a very dady thing for someone who cares about IT working for a nh is coming, okay? Once you have a huge customers who love your product, that will be very easy to expand from there. Let me give you the example of angolan.

Well, today I going is a very broad, large, like full featured search engine, right? But when we started to be nest, IT was just a glorified auto complete, very minimal. But that product was so much Better than whatever else was available at the time that some people cared enough about IT.

And when we when we started demonstrating IT, we didn't have merge 读书。 Actually, I still remember our first customer that first meeting when I demo using a common line to index content, we didn't have any A P， I, K. On yet.

We didn't have any, I mean, you, I yet, just a common line. And then a very simple way page to shoot search that was enough to close the two thousand thousand A S contract. okay? You have built something you can show users.

So the next step is to talk with users. I knows that many depth, many take of us actually in tot. And they don't like speaking with users or with customers like they don't like sales.

Well, you might realize, but as you are being in a death tool, you have a huge advantage. You know your audience, you are you o you are dever yourself. You speak the language of your customers.

You are uniquely qualified to understand them. So you see it's actually would be way or anything. Please don't wait for the product to be ready.

You want to learn A C, P right away if you are solving a new problem, and people will be ready to pay for. But how do you find these users? Two things, out witch and lucies.

Let's speak about out witch quickly. This week. Founders often plan to have a bottom of the adoption model where individual developers are going to come in down twice the product and convert, oh, maybe becomes a champions and bigger companies.

Well, I have some bad news for you. Nobody knows you yet. So to get there, first need to get people to know about you.

And there is another way to do that, then out watching and finding them yourself. But you need to be smart about IT. Start with your network, your press colleagues of class mates, and then you expand from.

There are friends of friends. You can live linking. You can find like specific cause over there, but please, please personalize your messages like, don't cooky what you can see out there. Send by marketers, actually your developer, i'm sure you hate these messages, so don't do the same.

Just ask yourself, would you be excited to open your message, then ask your founder, ask also live a friend, would they be excited to upon your method? And then when your email and then would be act on IT, well, you simply eat rate until they do. Now let's speak about launching.

You can launch many times, and that way you should. Do here to get more attention. The best plays to launch for their tools, I would argue, is happy.

News I can use is the community of intellect crary curious people. Many of them are developers. And there is that section in the company could show H.

M, that the perfect place to launch and get your feedback on what you are believing again, like when you wear writing like kind of these outwit messages, please don't do marketing, don't try to sell your products, just explain planning and simply what's new and interesting in what you are building that will be very engaging for the community and that share comments to you, engage with these comments, don't let them, and answered I to engage with people asking questions. Some of them are going to be haters and be hood. Well, you should engage with them too.

But what is not to convince them, it's just to convince any other leaders we are going to heed how you, he lied to these years. Just be classic here. One of the great example here is segment, how they simply did an experiment by a launching them next idea on high news, and they brew up, they got hundreds of invoice and so many comments that they knew they were onto something.

And that becomes the idea. Another example is ola ama helped companies run l ams on their own machines. Well, it's started as a comment on the comment on the noser post on high can use.

And just from that, I got a few people interested, and even inside of shortly after they lose their product, IT completely took tough on high ones. I got like, again, hundreds of her words, and then they will do something and they didn't stop there. They launch again and again every few months, each time they like new things to share and that continue to feel their growth, uh, creating a lot of excitement in the community.

Hacked your news, this place. But that thing I wanted to touch a point here in that section is you should do things that don't see you again. You are here to learn as from your users, what they like or not about your product.

A great example of the teams that did seems and don't scale is tribe. Early on the strike team, we are going to their customers, small startups, and they held them implement their own product sitting inside by side in front of the computer coding with them. That was not only a great way to make sure that users that are going to be satisfying to be actually excited, but I think but IT was also a great way together, invaluable for feedback, very early forms their users.

So it's complete OK to do things that esn scale if they lead to faster learnings. okay. And here, a few mistakes to avoid.

So you remember, the goal is to go fast and great fast. So please don't choose a text tag because it's cool, because you want to run IT. No, you should do IT because of your expertise visit that you can iterate as fast as possible.

All our mistakes we are, we touch, not talking with users. Please get that feedback as fast as you can. We also means the avoiding overbuilding before getting that feedback may be a small thing here.

Be careful about that feedback. Be careful about not me. So understanding some of the developer of feedback, for example, they may share, may tell you or you shouldn't beat your polar, you don't care.

You just want to know if they would use your product to lot and why, or they could tell you, you know what, I could build this in a week. I don't need IT. Well, guess what? You just build IT yourself in a week.

You you should not find about that. You just move on and get new input. The product last mistake some teams do is hiring too early, very early on.

You don't even know if your product uh, is providing value. Please don't hire anyone until you have convinced ed yourself you are working on the right idea. Let's now talk about go to market.

In the first question you should ask yourself about go to market is your business model. What basically should you be open source. Open source has become one of the main good market at four f tools.

Any different key works as a business model as many companies are over like dataworks, eli cashing club, great club or awesome, great companies. I would argue that you must be or if you are working on things like a library of homework, that's what's today, expecting the market deers. We don't implement on top of a new homework that is not open source.

You should also be open source. If you are dealing with data as a sensitive data, a new database would need to be a pencil. Oh, N, E, P, R, connecting with sensitive data in the sir ham, or maybe a near needs to be a en source.

You could be open sore, but probably don't have to be en source if you are working on those of things like A P I, or maybe an application like a bus cracker. Maybe in that case, it's not as important but still could be a your deformation open roles as quite a lot of benefits your customers are developers. Well, as you know, already developers prefer working with open sls tools.

It's also going to help you create more awareness ness through the IT can be also your different like you can be the open service version of x think out post super great approach. Speaking of their bite, you could also receive contribution from the community in the case of your but I would be careful on that one. Most cases, you couldn't count them on them because the quality of the contribution may not be up to your standard, and then you have to deal with the contributors.

And that could be more actually more difficult than you think. But in most cases, I would say that you should not content them. Good contributions actually pretty are and last benefit, but not the least is the trust.

And it's actually very important for large enterprises in the enterprise. Real takes example of meet plan, for example, with building a abou hr. The fact that they are open stores, I would argue, publish, help them shot, turn the sale cycle with enterprises by a year more.

There is one thing you need to think about ever when you are creating an a source project may be not have to start, but pretty soon it's moniz ation. If you cannot moniz your product, you don't have a company at some point. Even if it's later in the life of the company, you will need to earn money.

There are many reasons out there, but it's good to look at the common approaches. Some of the common ones are hosting adding a cloud offer so your customers may use your opening source, he said, posted, but they could also pay for you to host IT so that they don't have to maintain IT. Sometimes clod version also have like some additional features like team management and those someone is being open call in that case, usually have A A special offer to be under Price offer that have advanced features that are not in the usual version.

Lions common features are typically the ones that are going to be necessary for the prises like S A soul of clogs, disaster recovery, S, L, S, these kind of features. And finally, some open source companies charge for support and services. But I would actually discrete you to follow that monitor pass, because your incentive becomes to build the most complex possible.

So you can charge for supporting services, which is bad incentive. And if your product need too much support, then people are going to turn. And not, he knows people.

And what if you are not up as well? For these death tools are usually following one of two approaches. These are are usage based, like A, P, S, I go yeah to use stripe. And so they charge based on the usage uh, the customers do uh possible. They have a volume discount or enterprise fix options that are on the side.

And the single approach is I think some specific plans typically like a good, Better, best apart good is going to be the same of option um that's going to solve the problem of death for a hallow Price for the Better options usually are going to be interacting with ending manager is we are going to care about that's going to enable teams still can be self serve. And the best option is higher. You are talking to the co.

It's all about security or gs disaster recovery, S L A. And in that case, it's usually a says that approach. Let's speak about sells. So as a quick reminder from what I was saying earlier, don't fear out, witch, even if most death tools are going to end up being bottom up and we lie on inbound, no one's knows you at first.

And so you need to start without which that period of founder al rich is how you start as you get more well known that when you get inbound, usually are going to have, at that point more self leads. And then eventually you end up iring your sales team to sell to bigger icons enterprise. So should you have ourself team?

Well, I would argue that you should wait as long as you can start by selling yourself. Don't delegate away the tooth is that you are the only person in the company today capable of selling your own product. If you at the founder cannot sell your product, nobody else can.

As a whole sum, I would wait to get to about one million A R before hiring your first sales person, and even then quite to hire your people. Work technical, at least understand developers, because developers not hate speaking with sales people, actually a number our first sales people that they need to change title on linking. And instead of like being cells or I can't exist, they called themselves product specialists.

And the nice thing here is that they actually had to own that status, that to understand the product in timely. So IT was kind of like a great incentive for them to be Better at product and a great way for them to open doors with developers who don't want to speak with cells. And there are example, post hog.

Did you know the team steel is also, they are sales leader, not a business leader. Their sales leader is also the C. T. U he sees sales as an in problem, which is a no so approach for their tool.

If you want to learn more about sales, you should also watch pete talk about enterprise sales or start up school video, so is different to sell death tool from sitting a news product? Yes, IT is you should didn't build a sales deck. The local is like, don't want to waste their time reading with sales that you should show, not tell, just do demonstrations.

I think that the best are sales apple you can do for the tools at algoa. We didn't have a sales deck before we got to ten me on A R. So even ourselves, like proper account exec, we are just showing demonstrations they were selling and that was working so much Better for everyone.

You should lean completely in the tech aspect of your product. For example, I have going we said a lot to e commerce players. And what realized is that when the buyer is not technically and want a turkey solution, we always, not always, but most of the time actually is the deal.

But when the buyer is technical or if take as a lot of influence in the decision, we win more often. So we realize by leaning into that aspect for product, we are nap. I we are not like a turkey solution.

Well, when did up like finding or I C P. And having metal success selling. okay. And the last thing that is also too in within death tool that most of them are a bottom up. What that means this is that a lot of people are going to self serve even within enterprises.

So instead of going a top down, just pitching your product directly, cells check if some of the enterprise employees are only using the product and they should then focus on helping management understand what the product is already doing for them and the home, much more IT could help them if they were. We have to pay for ball like new seeds or features or active usage. okay.

So the last uh topick I wanted to a touch in this video is developer marketing. Your goal is to create a lot of awareness for your product because you want in down. The first thing you should do is find your community so could be back on news or some or some distance out there that he is related to what you're walking on.

But be careful, your goal is not to sell your product just yet here. Just be helpful. Just estimate yourself as the expert.

And of course, people are going to start knowing you and trust your brand, your company, that will help you done the line. Uh, people think about you when you have the problem launch often. That's very important.

We discuss about the example of before, uh, but that should also be done for the life of your company. Some companies even do lunch weeks, and i've been a very successful for that. This works great for sorts.

For example, super base is doing that once a quarter, burning all the animals in one week a quarter to make as much noise. I say all the things you can do make your documentation of first class citizen documentation is often left of start for delivery. A prefer coding.

That comment that I would ask you that the city should never be to be part of the product document is marketing. That's how people are going to interact with your product. That's actually the first place people are going after your home page and maybe the pricing page stripe and also your death tools ever really raised expectations there.

People person have become very demanding that something you should invest in at a golia. We decided that the feature was not ever done until the duck was done, and the city should be written by developers. good. Even when we chose the team at some point, uh, we creating that documentation team that was not to write the documentation for developers, they were here to help developers write Better documentation.

For example, they build the tool where IT would automatically insert A P I keys for people who be logged in, so that when they copy, could IT would walk right away at those books, uh, which means s the condition everyone more useful for our customers. Speaking of things that depth should do, how would include ort there? You should make support a huge part of your marketing.

Your users are liberals. S White developers hate speaking to this livery. One support people who don't know anything about the product they are supporting, engineers should do support.

And that's going to to lead to two great things. One, your users are developers. And when they interact with developers, we speak their language, are going to be when more satisfied.

Actually, I have examples where someone would trust to support. And then the developers, like the person doing the support, was able to actually fix the bug and the in production in twenty minutes. And so that kind of like blew the customer away like they were so impressed, so excited about the product.

And too, it's going to help your own diverse, much Better understand your customers need because they are going to face the pain as they are supporting them. And that's a great way for them to be the Better. Taking the example, lie again.

We actually waited hundred, hundred hundreds of police before we build a delegated to team, and then we actually asked to one best engineering manager to build that team. Support was never a second class design for us. So deeper peers are doing documentation.

Reports are doing support. All of that is your marketing. Should you even hire marketing team? Or when should you do that? Well, similarly to sales, I think founder should lead marketing for a long time, may be forever actually, to be even a later stage.

I don't know a single death tool company that is happy with this when they have a trade marketing background. why? Because traditional marketers, they don't tally understand the personal details, have please don't insult that, you know that because you are right. Well, only on we were actually asking every single engineer in the team to do what we called a marketing hack. Every single months, a hack could be writing a simple breakfast, or could be speaking at to meet up or could be building new search.

Our best marketing and was done by the I wouldn't code you as you build your marketing team to make sure that you have a lot of developer, your marketing team because the other ones who know how to speak to your auto, you may ask, what about that advocates? Well, I love the idea because that means having developers in your marketing team. But it's also a very ill defined hole.

We found a lot of very fuzzy expectations. So it's difficult to measure, difficult build accountability for. So my recommendation here would be to wait and first leverage your existing engineering team and ask them to do some form of marketing or their advocates.

And probably a little later, IT starts to make sense to hire a dedicated device. Cates, even then think about hiring from within of from your committee, for example, and source could be contributors or active, helpful disco members. These are the best educates candidates to join you.

So let's rap up and he cup what you said here. And now please don't wait for the perfect idea, because as you are going to build your product, even if it's not the perfect idea, you do the raps you learn and that will lead you to the best idea. If you don't start, you'll never get anywhere really quickly.

Please, please don't over engineer. You need to have that very fast iteration pace where you can learn from your users as you are building coppy quick and dirty is completely OK every stage, and then you effect, or what actually matters, spend time with your users. Miki is know where you're going to learn. If you don't speak with your users, learn from them, do things that don't care, go to their office, spend time with them, launch early maps in many of our videos where we say there is no shame in launching early.

People won't remember who we think doesn't work out, but you could get available feedback that will lead you to work on the right product to build the right, think that people need, as you are being your death to you should consider open source as you go to market, observe the market, observe other players how they build their product and figure out if open sources, the writing for you. But don't this music think about IT before making a choice. Don't forget, you are the best sales person for your product.

No one else is going to know your product or your audience as well as you do. Learning to sell is very easier than learning how to speak with the developer. And finally, you are also the best marketing person for your product.

You know, the along gradually dies, you know, to speak with them. You know where spending this at times the community, you know how to access this. Uh, audience, okay, that wrapped up for me today. Thanks for watching the video. And if you are building a death tool, don't forget you can apply to, I see any time.