I've been thinking about writing a post on the role of a CTO in a startup for a while. However, I always told myself it wasn't interesting, that I should talk about it when the company was much bigger, when we were a unicorn, etc...
And so I kept delaying this post for months, then years...
However, over the years, I realized that the journey could already help others.
I have benefited and still benefit from the feedback and experiences of many other tech leaders and entrepreneurs who are at more advanced stages.
But while I read and am very inspired by successful entrepreneurs or product leaders like Jean Baptiste Rudelle, Reid Hoffman, Marissa Mayer, Marty Cagan, I also know that when you start a company, these people can't help you. Their daily life has nothing to do with mine.
Moreover, while there are of course hundreds of thousands of tech leaders who are CTOs of consulting companies, development directors, etc., with engineering teams larger than the entire size of Malt, I think the job of CTO in a growing startup context has some specificities.
And that's how I decided to talk about the CTO role, from startup creation to its transition to scale-up.
To set the context, I am CTO and co-founder of Malt, a marketplace allowing freelancers to work with companies of all sizes, from SMEs to large CAC40 groups and now in several European countries with offices in France, Spain, and Germany.
The company has almost doubled in size every year (except 2020...), going from 3 people in 2013 to 200 in 2021. The product team currently has 50 people.
Over the next 18 months, we plan to double the size of the product team to about a hundred people and accelerate our expansion in Europe, making us enter the status of scale-ups.
Let's start from the beginning.
CTO, boss of myself, in a team of 1 (2013-2014)
Do you really need a CTO to start a startup?
Malt started with a meeting in 2012. At that time, we were 3: Jean Baptiste Lemée, Vincent Huguet, and myself. The association would therefore count 2 "CTOs" and one CEO.
I often have discussions about the best teams to start a startup. Can you start alone? And if not, what is the best team composition? How should roles be distributed, etc...
It would be a much too long article if I tried to answer all these topics, but I'll focus on the CTO role.
I won't beat around the bush: from my point of view, the CTO role is crucial for a company that wants to build a product.
I don't know any startup that has passed the 3/5 year mark without a CTO.
I've heard from investors that it's even a serious red flag in a startup file to come without a technical co-founder.
You can of course start an idea without a CTO. You can make first versions, and even do a first funding round, with a product made by an agency, or made with no-code tools.
But at some point, the CTO is necessary for the company to develop. The CTO will bring the vision and technological governance necessary for the future.
On the other hand, and this is what I'll try to describe in this post, the typical portrait of the CTO at this stage is not the same as the CTO in a company of 500+ people.
CTO is above all being the first developer
As co-CTO in the first months, I was mainly boss of myself. I was above all a developer and my main activity with Jean-Baptiste was to code the Malt website.
Product definition was done among the 3 founders.
It was a mix of dev, choices between buy or build (the famous buy or build) but with makeshift solutions. There's no bigger challenge than speed at this moment.
Of course, I didn't want to build a shaky house of cards, but you had to be pragmatic and ensure very good execution speed.
Full stack CTO
At this stage, you're involved in many subjects you don't know: SEO, data analysis, data engineering (poor man's version), data science (even poorer version), system administration, QA, PM, UX Designer, front-end, back-end, etc... You define the product and write it right away.
But it doesn't stop at the product, it's also IT support, customer support, finance, etc...
Of course, I wasn't good at everything. And today they still tease me about screenshots of my old UX work. Anyway, the CTO is the person on whom you can then stick all the "Historical" baggage, with a capital H, and you have to live with it :)
By taking on the entrepreneur hat, you have to remember a phrase: "If you're not ashamed of your product, it means you released it too late." (Reid Hoffman, co-founder of LinkedIn).
The typical profile
The typical profile at this phase is a person who gets their hands dirty and moves fast. It's also a person who likes to learn new things and is okay with being constantly out of their comfort zone.
By nature, a good engineer is rational, rigorous, methodical. But be careful, dogmatism is an enemy at this stage. You have to accept a dose of creative naivety and the very many compromises with what you could have done in much larger teams. At this stage, you shouldn't seek to be a person who will structure a team. You shouldn't spend time on process optimization or organization.
Compensation
Salary-wise, the CTO at this phase shouldn't expect nabob compensation ^^ You'd earn better elsewhere.
This is the major difficulty in finding a CTO when you're not one yourself. If I exclude young CTOs, any experienced person would be better paid elsewhere than being CTO of your startup. Personally, I earned much better as a freelancer at that time (>50%) and working less.
Associating the CTO with equity is important to engage them for the long term. The CTO who joins at startup expects 30/45% of shares. I won't dwell on the criteria of how to determine this, but there are sources on the internet to help you.
For my part, being an associate was necessary, but I never told myself that it would compensate for what I was losing. Of course, when you start, you want the company to work, but to be 100% sure that the company will one day be a complete success, we know how much it's a bet and that very few people make it.
On my side, I accepted earning less because I found it very enriching. I always loved creating my own product, making it grow, learning new trades, and being my own boss. The freedom I had and the impact I brought motivated me in this adventure beyond the financial aspect.
CTO, squad leader in a startup from 1 to 20: (2014-2016)
(product team from 1 to 10)
There you are, the first hires in the product team. Your CPTO (Chief Product and Technology Officer) role grows. Product definition is still done among founders and you're still mostly hands-on more than 90% of the time, but you have another responsibility: creating a team.
This is a phase that is actually pivotal in more ways than one, but I didn't know it yet.
Learning to "fire yourself"
This will be one of your mantras later, how to constantly fire yourself?
Now, you need to hire colleagues to do certain jobs and do them better than you. No more PM, UX, Front-end roles, etc...
I have the impression that some CTOs get stuck at this stage because they can't hire better than themselves. Is it related to a feeling of insecurity? A way to want to keep control? No idea. But it will be very hard for the future...
The CTO, still a developer
At that time I was 90% developer. 2015 corresponds to my peak activity in terms of commits on Github (see below). The product team was about ten people at the end of this period.
In a team of this size, autonomy is great, the methods used are very light, and the famous "time to market" (the ability to quickly go from idea to users) must remain very short. There are no estimates, no sprints, scrum master, QA, OKRs, and so on. Decisions were quick with few people to convince each time. Change management was "simple" (and yet we sometimes tripped up anyway).
A recruiting career that starts
Recruiting is 90% of the company's success and that's why I spent a phenomenal amount of time on it.
In total, I must have seen almost 1,000 CVs. I conducted 300 first video interviews, between 100 and 150 physical interviews (also counting for other roles in the company even if I wasn't the main recruiter in those cases).
Recruiting also means making choices about team organization, the culture you're going to create, the fundamental values you're going to instill. This translates into seniority, autonomy, adaptability, entrepreneurial spirit of the people sought.
I made a structuring choice at the time, I chose during recruitment to only take seniors, completely autonomous people. I didn't want to do management and manage juniors or coach intermediates. The organization remained light, communication simple, autonomy great.
This philosophy allowed me to continue developing on Malt for quite a while, but beyond that, it seems to me that it's a success factor and part of our growth.
Your CPO/CTO/founder role requires more time.
Beyond your role in the product team, your role in the company also takes on a bit more importance. You have more interactions with more people outside your skills, trades appear like support, sales function, marketing, and you have a business enabler role (sorry I couldn't find a satisfactory French translation). You carry the "voice of technology and product", it's up to you to identify when it can serve the company's needs, when it can bring answers to your customers, but also when you need to know how to make choices (and say no).
I specify of course that this role was not done solo. As since creation, it's the trio of founders who play the role of full-stack entrepreneur at all levels.
CTO, guild leader in a company from 20 to 50 (2017-2018)
(product team from 10 to 30)
More complex interactions
In 2017-2018, I spent a large part of my time recruiting. And by doing this, I passed the fateful threshold of 10 people in the product team.
This threshold of 10 starts to change how a team functions. Interactions are more numerous, communication becomes more complex, and small groups start to form to remain effective.
Brooks' law perfectly illustrates this growing complexity in group interactions past a certain threshold:
It was at this moment that I was delighted with the team already in place, on which I could rely. The fact of having recruited many seniors limited the need for human management. Some leaders naturally emerged and the team was able to continue functioning without heavy management for another moment.
My model was that of the previous company I had co-founded, Lateral-Thoughts: sociocracy and governance by circle.
The CTO role at this stage is very team-centered, very present at all levels, the CTO is there to compensate for the absence of formal structure.
And this is not abnormal, it was still too early to bring more structure, and at the same time, we had to start working on subjects like alignment and long-term vision.
Without being a pure people manager, I had to greatly improve the recruitment process and work on subjects like career path or salary grids. However, I didn't set up a formal process to see each person in the team. I was counting at that time on the fact that discussion should happen naturally and that everyone was big enough and close enough to me to talk to me about subjects that bothered them. Spoiler, it worked, partially. At this stage, it was still possible, we saw each other a lot and communication was very open. However, during the next phase, this principle no longer worked, especially in the context of the COVID crisis.
A developer role that decreases
As CTO, I'd say my share of time spent developing must have decreased. I'd naively say around 60%.
The organizer role took more place.
It was at this moment that I started defining myself as the GO (Gentil Organisateur - Kind Organizer) of the team. The person in charge of putting oil in the gears.
It was still a schizophrenic role: developer, coach, manager, scrum master.
If pure development decreased, it remained important and key on orientations and technological choices. The activity changed in nature, but I'd say this developer role remained major for the company's success. At this stage, I was still able to bring significant changes to the application in a few days.
To illustrate with numbers, here's a graph representing my Github activity from 2012 until today, which well illustrates the decrease in activity in 2017-2018 (and the 2015 peak):
A permanent ambiguity
As mentioned above, as CTO (and founder), you constantly change hats, several times in the same meeting.
You're still a "full stack CTO" but in the meantime, you've recruited, and these permanent hat changes can bring some confusion for company members who arrive.
It can be difficult for everyone to know when the founder is speaking on a subject as a "participant", for example in a brainstorming session, and when the CTO is speaking as a technical leader, to express a direction to take or as CPO to establish product strategy.
I had many discussions with Malt members to clarify my positions, assume this CTO role that was until then mainly a label I didn't claim, improve my communication. And I greatly thank all the people who at that time pushed me, made me think and grow.
It was necessary just before the next phase because I would have to make stronger choices about what I wanted to do or not.
CTO, tribe leader in a startup from 50 to 200 (2018-2021)
(product team from 30 to 50)
The 2018-2021 period is special in more ways than one. Between 2018 and 2020, the company doubled in size each year. 2020 was marked by COVID with different management to implement and turnover that took off. In the same period, the company progressed several notches at all levels.
A team that needs to start being structured
I mentioned it above, the complexity of communication when you increase the number of people involved grows exponentially, and it was during this period that we had to pass many milestones.
As CTO, it was time to bring structure so as not to degrade execution speed. If too strong structuring could have slowed down Malt at the beginning, it became necessary to then scale.
This is the moment to realize that a product team requires impeccable coordination and animation. In a tech startup like Malt, at this stage of development, the size of the product team should take a ratio between 25 and 50% of total staff (ratio that decreases with growth thereafter). It's therefore one of the largest teams. Each squad/feature team/impact team/tribe, call them what you want, corresponds to collaborations of several different trades. They're not just mercenaries working each on their side. Each person, each decision influences another person in the group in the short or medium term. Moreover, almost all people in the company rely on the product or tools made by this team.
Technical or business debt is collective. But if we also see it positively, each improvement, each good practice substantially improves the entire team and company.
In short, it's a complex mechanism and, fortunately, there are many companies that share their best practices. As CTO, you'll need to apply constant watch on these subjects, the same way you were used to doing on more technical subjects. I also propose a list of resources at the bottom of this post.
Time for choices
I mentioned it at the very beginning, the CTO's goal is to "fire yourself" with startup growth. And the transition to 100 in the company accelerated the movement.
At this stage, it was time for choices. As CTO, I could no longer develop, organize, establish strategy, recruit, coach, etc... Or do everything badly.
In the rest of this paragraph, I'll mention the personal choices I made. Each CTO (and this is even more true for a founder) will make different choices. There's no right or wrong solution, you also have to work with your strengths, weaknesses, and interests.
First decision, I almost stopped all development. OK sometimes it comes back to me, but it's anecdotal. I always need to understand what's being done, but it's mainly out of curiosity. I should never be in the blocking path for development.
Then, I always considered myself an individual contributor, not a manager. So afterward, I brought in a VP of engineering to take care of management and processes.
We brought in a CPO on product strategy in 2018. However, this role disappeared in 2020, and I have a role that de facto became that of CPTO again, but with a team of super strong product leads to help me and with whom I'm delighted to work daily.
Thanks to these leads, I remain very involved at the governance level on subjects that are and will remain crucial for Malt's scaling.
They accompany me in all key areas: security, architecture, data, cloudops, product strategy, design, etc... I spend a lot of time with them on strategy and objective definition, problems to solve, but then they're autonomous on how to do it, and I help by putting oil in the gears when necessary.
And what's next? The transition to scale-up
As I indicated in the introduction, Malt is starting to enter the restricted world of scale-ups. The next objective to respond to Malt's growth and its expansion will be to continue investing in the product team. It could double in the next 18 months and reach about a hundred people.
Scaling this team will require enormous work. A large part of this work rests on the CPTO and the product leads team I mentioned previously.
The stakes are high: recruitment speed, quality of new arrivals' training, team autonomy, execution speed, product strategy, user research, OKRs, technology scaling.
The CPTO role will undoubtedly be very strategic. I'm going to work on OKR deployment at each level and on monitoring and coaching the product leads team to respond to these subjects. At least that's what I imagine, but the past has shown me that the most important thing will be adaptability.
This might be the subject of a new post in 18 months.
And if you want to be part of the adventure, we're constantly recruiting throughout France. Contact us.
Interesting resources as CPTO:
- Inspired and Empowered by Marty Cagan, the two books I so regret not having read several years ago. You must read them :)
- Turn the ship around by David Marquet, on leadership and how to create leader/leader relationships
- The future of management by Gary Hamel, a must-read on innovation and managerial structures
- Measure what matters by John Doerr on using OKRs
- On m'avait dit que c'était impossible by Jean Baptiste Rudelle on entrepreneurial spirit (only available in French, sorry)
- https://randsinrepose.com/welcome-to-rands-leadership-slack/ for tech leaders worldwide (14k members)
