Wanting to rebuild everything from scratch in the name of sovereignty is a false good idea.
Building a European cloud ex nihilo to compete with AWS or Azure, and more generally hoping to build an “Airbus of [insert whatever is trendy here]” in tech, means hoping to take a place in a market where incumbents are firmly entrenched: they have millions of customers and years of accumulated network effects, in a market where interoperability is still a distant utopia.
Public procurement to build a customer portfolio that offers visibility? Hard to count on, given that the political reflex in recent years has been to spread money thinly: we subsidise start-ups, forget the established players, and end up launching an in-house project in a ministry to compete with the very start-ups we subsidised.
So how do we do it?
Network-effect economics distinguishes three other distinct approaches:
- Competition between solo players: it splits the market into segments where each gets its share. And in the end, the giants already in place win.
- Merger: with network effects, 1+1 does not equal 2. Each party loses its differentiation, weaknesses are pooled, and time is lost focusing on integrating the merged entities.
- Cooperation: we interconnect, we become interoperable, customer bases move smoothly from one to the other, or even use both at once. We fill the infrastructures and maintain strong robustness. Each keeps developing its own identity and focuses on making its product the best it can be, simply offering it to a combined customer base.
That is exactly what France-IX does in peering, a network economy if ever there was one. We interconnect IXPs and give access to all the interconnected data centres from a single port, across France and several countries.
European technological sovereignty will not come from a big centralised project. It comes from players who cooperate skilfully and share their infrastructures to create a common ecosystem.
What if the future of European tech were extreme interoperability?
Multipass
This news was first published on LinkedIn. To comment or share it, head to the original post. → See the post