<aside> 💡
Headline lessons
M-Files started out in 2002 from noticing a recurring pain in everyday work. Antti had seen how hard it was for companies to find, manage, and trust their files, including in his father’s business, and he carried a clear product intuition that “this should work better.” On their journey to building M-Files, the team became experts in the slow but steady way: by building, selling, delivering, and living inside the problem for years.
Antti founded M-Files and then recruited three cofounders early, who became the first leaders of sales, services/implementation, and additional product development while he stayed as the product person. The founding setup worked because the team covered the whole loop, not just the product. Once the first version shipped, this division became their way of operating. Product could go deep without the company neglecting sales and delivery, and customers could be sold and delivered to without the product turning into a sales-only shell.
On a general note regarding centaur-building, Antti mentioned that when building a team, you want a useful kind of internal tension where you have overlapping priorities. You want one person to be pushing for speed and momentum, while another person protects quality in problems that genuinely take time to solve well. The main goal is to build a founding dynamic where speed and depth keep each other honest, but also balance the other out so as not to drift into perfectionism or ship a shallow product.
Before M-Files, Antti had already lived through what “too small” feels like. His earlier product, M-Color, served a niche so narrow that a single city in Finland might only have a handful of potential customers. That experience gave him a strong motivation to choose a bigger opportunity next time.
With M-Files, the underlying problem was clearly big enough. Antti saw document and information management as a category that would only grow in importance, because the amount of content inside companies was increasing every year. The mistake, in hindsight, was how they translated that into an initial market choice. They started without enough deliberate market study, and they were not deep domain experts yet, which made it easy to define the market in a vague way.
That is how they ended up over-correcting. Instead of picking a focused wedge, they framed M-Files as something for almost everyone, across all industries and company sizes. Looking back, Antti thinks that was too wide to be a good early entry point. If he were doing it again, he would still keep the market big, but he would narrow the wedge by choosing an industry vertical and sizing it globally. He also points out a very Finnish trap here. If you only look at Finland, almost any vertical looks too small, so founders are tempted to say “we serve everyone.” His view is that the right lens is global, because a “narrow” vertical like manufacturing becomes huge at world scale. The goal is not to stay tiny, but to stay focused enough that you can win a clear segment before expanding outward.
M-Files’ early build is one of those rare cases where a long R&D phase actually happened and still led to a centaur-scale company. They set out to build M-Files in 2002, with a plan to get the product working within a year. In reality, they spent roughly three years building before shipping which was seen as a waterfall-style effort where they tried to define what the product should do, make architecture decisions, and build a deep foundation before going to market.
Two things made this feasible.
First, they had a funding through Antti’s previous company. M-Color generated enough recurring revenue to finance a small team, roughly five people, to work on M-Files for years without outside capital. That effectively acted like their seed round. Instead of raising money to buy time, they used a smaller business to create time.
Second, they used that time to build heavy foundations, including building their own virtual file system driver, which required time to make work. This period came out of necessity, as the team believed they needed time to get a product of this nature into users hands.
At the same time, Antti is very clear that this is not a play most founders should copy today. His advice now is almost the opposite: get an MVP or proof of concept into real customer hands within months, because modern tooling and AI make it faster to test reality. The goal is not “hide for three years,” but to be honest about what kind of foundation your category truly requires, and then not using “deep tech” as a story to avoid learning from customers.
Once M-Files had a first version you could deploy, the company moved from building in a room to learning in the real world. During the first three years they did not work with pilot customers. They built what they believed the product should be, shipped a first version, and then started selling.