A major part of the infrastructure supporting the digital revolution is designated as “the Cloud”. Far from being solely virtual, it is composed of brick-and-mortar servers and hard drives, hosted in massive datacenters and consuming huge amounts of power (no, these clouds don’t fly). The largest player in this infrastructure space is Amazon Web Services, weighting an estimated $6 Billions of revenues in 2015 — more than its 4 largest competitors combined. It has two European presences, one in Ireland and another in Germany. France, in its Colbertist tradition — or was it arrogance ? — decided that this could not happen freely, and had to be shaped by a government-mandated effort. This has made sovereign cloud a top-down failure.

In 2009, the French government unlocked €150M to build a France-based cloud, a sizable sum for a budget running a constant deficit. Orange, Thalès and Dassault Systems were listed as co-investors. But as a sign of things to come, Dassault Systems pulled back, and we ended up with 2 projects instead of one:

• Cloudwatt (Orange + Thalès)
• Numergy (SFR + Bull — now Atos)

Each project received €75M each of taxpayers’ money.

To make a long story short, we are now in 2015 and these companies (despite TV campaigns as costly as irrelevant for a B2B service) are virtually non existent in the market: Cloudwatt is reportedly making €2M of revenues per year, and Numergy €6M of revenues per year. This falls short of the €400M promised by the business plan.

While so many companies struggle to raise money, this innovation process raises eyebrows:

• First question: if €75M was enough to start anything, why grant €150M in the first place ?
• Second: why start such a projet, when the 3rd web hoster on the planet is French ? Yes, there was already an extensive Cloud infrastructure in France, and we could have helped the best become global faster.
• Third: what is Europe doing ? Such subsidies distort competition and are not supposed to happen.

Truth be told, I can’t help thinking that insane friendship between politicians and VPs at these companies was involved. But I don’t want to even find out — for fear of shame of being French.

So let’s take the best out of this experience: I am not talking about cheap second-hand server hardware, but rather learning. What went wrong ?

Simply, innovation does not happen by just deciding it. Money does not buy an innovative environment. I will remember that:

• Numergy and Cloudwatt took years to just catch-up with Amazon or OVH technically. Forget about being better as a late contender.
• “Fish rots head first”: the project was ill-conceived, but the executives that were hired where not any better. There is no product differentiation, just expensive and inefficient TV campaigns, no presence on the field…
• Telcos: I hear a lot of telcos willing to deploy their own Cloud infrastructure. Fair enough, but look at Numergy and Cloudwatt as possible outcomes of these initiatives. There has to be another way.