#126 - Using the cloud sovereignly – when hyperscalers make sense and when they don't

Joubin Rahimi

Great to have you back for a new episode of insights! My name is Joubin, Joubin Rahimi, and joining me today: Marc Achsnich, a Fellow of TIMETOACT GROUP and, within synaigy, the person driving Managed Services and the topic of Cloud and Sovereign Cloud. Hi Marc.

Marc Achsnich

Hi Joubin.

Joubin Rahimi

And the nice thing today is, I'm more the unstructured type, Marc the structured one, and he asked me. Did you read the questions? No. But we still enjoy working together, and I did skim through them again. And we're keeping the intro light.

Marc Achsnich

I'm keen to see.

Joubin Rahimi

We've already covered quite a lot of ground together on the topic of cloud. And we had AWS, I think we've now had 15 years of AWS partnership, but we've moved part of it away from AWS. And you were significantly involved in that, it was one of the first activities. But how did you come to the topic of cloud and sovereignty?

Marc Achsnich

Yes, that's quite exciting. I think sovereignty wasn't really our core focus at the start. It was more a topic that came from our customers asking us: you're on AWS, that's our main competitor. Why do we actually have to pay you, or AWS, money just to be allowed to sell. And also at synaigy one of our guiding principles has always been that we stand for diversity in retail, or also that we say, if we can simply do something from Europe, then we should just try it. And that became a starting point where we thought, there's something to this, and we went looking to see whether there are simply alternatives that just aren't a hyperscaler.

Joubin Rahimi

Und wie war das damals? Vor wie vielen Jahren?

Marc Achsnich

I think that must have been five or six years ago, when we slowly started out and thought about: what is it we really need? What use cases do we have that need covering? And then took it a level deeper, because in the end the question is, Kubernetes for example, in theory everyone has it, but then the question is: what do you do with it? And what level do you actually need? And then you already noticed, and that's where the wheat gets separated from the chaff.

Joubin Rahimi

And, how exciting is it, we've also been an OVH partner since that time on one hand, and now we're also an award winner, I believe, at that point. I'd say Marc is also an MVP for OVHcloud. Why? Maybe you fell a little in love too.

Marc Achsnich

Yes, maybe that's one of the points too, because I said, because we love what we do. And maybe that's part of it, because I just enjoy reporting on what's ultimately possible with OVH or with Europe, because when you talk to a lot of people, it's always "that doesn't work, this doesn't work." And no matter what you have to do, it has to be a hyperscaler – and that's simply not the right answer any more these days.

Joubin Rahimi

And why is it so important that it's no longer a hyperscaler?

Marc Achsnich

I don't actually know whether it's necessarily that important or whether that's the right answer, because I don't think there's a wrong cloud per se. Every cloud is a valid player in its own right, has its individual strengths, but I think you don't always have to reach for just one. And that's diversity in retail again - the variety that matters simply ensures that everyone moves a little out of their comfort zone, and something better emerges from that.

Joubin Rahimi

And with Trump now in power for the second time. I think the first time round I asked myself: which mail system could we use instead? And found there's hardly a German or European mail system that springs to mind immediately. There are of course niche products. And now with Trump in power again, the topic of sovereignty is back on the agenda, and noticeably more intense and emotional in the discussion. With customers too, or is that more our service-provider view on it?

Marc Achsnich

I actually believe it comes from both directions. We've also had various customer enquiries where suddenly many exclusion criteria appeared, stating that EU providers are simply set as a must-have, because it has simply led to a climate of fear. There's also the Bitkom study from last July, I think it was, which said that over 96%, I believe it was, also simply fear negative impacts due to the government, or genuinely see an impact on their business. And fear naturally drives that kind of behaviour. In the end, the only question is: is that justified? And is it wise to act driven by fear, or not.

Joubin Rahimi

Or not, right, or not. Right, usually or not, yes. And are there certain patterns you'd say companies should follow to make the right decision here?

Marc Achsnich

Well, it always depends on what data you have, or what your USP ultimately is, because there are, as I said, plenty of cases where I'd agree AWS can be the right solution, because not everyone has highly sensitive data. Whether I'd personally endorse, for example, the Bundeswehr deciding to hand over its internal IT to Google is perhaps a different matter worth evaluating separately, but on the other hand there are simply so many use cases where it makes complete sense. Or you might say, I'll accept certain trade-offs, because there's something called a protection needs analysis. That means, on that basis, you look at what data you have and where you need to take action. And then there are plenty of areas where encryption or MFA or other measures are sufficient, and then a hyperscaler is entirely fine, and you get to use the innovation it offers too.

Joubin Rahimi

Where would you see OVHcloud, where's the sweet spot there?

Marc Achsnich

OVH has the advantage of having been on the market for quite a while and, relative to other providers, having a very broad portfolio. That is, OVH sees itself in the range of around 100 services one can choose from – not at the detail level an AWS actually offers, but around 80% of the use cases a normal customer needs are quite achievable. And that's one of the advantages. In terms of OVH's history, they started as a company that simply produced servers. That was Octave. When he founded the company 20, 25 years ago, it was a small hardware outfit that just made servers. Then they built a private cloud out of that, and then built that into the public cloud. So compared to the hyperscalers, they simply have the ability to offer bare metal at unbeatably low prices. If you go to an Azure now and want bare metal, you'll pay something four-figure for what you could otherwise get for, say, 100. And there you have the option to simply combine everything together. And that's something very few manage at this level.

Joubin Rahimi

And towards data security, and you said if highly sensitive data is involved, that's a case where you'd rather shift the weighting towards OVH, right?

Marc Achsnich

Yes, the right answer there is also, it depends. In the end the problem is that sovereignty is simply far more than data, because that's a completely different driver of it. If you only look at data, the hyperscalers are better. That might surprise you now, but it's simply because they've been dealing for much longer with what a provider can do so they don't see the data, because it's also the case there that AWS doesn't act maliciously. They know there's the Cloud Act and in some cases they have to hand over data. So what do they have to do then? Encrypt it so they simply can't do anything with it. And that's why they're simply far better at this level of data sovereignty than an OVH or a STACKIT or Informaniak, because those only started dealing with it later due to regulatory requirements, since by law they aren't allowed to do it anyway.

Joubin Rahimi

And where is OVHcloud's sweet spot?

Marc Achsnich

On one hand in the pricing structure. That means we've done studies, you get something offered at 60% less than what you'd build with the hyperscaler. And then of course the question is, depending on how big you get, you reach a situation where you're burning through the equivalent of a detached house every year. Whoosh. Yes, and that's the question, I think people are reluctant to do that.

Joubin Rahimi

Regarding cloud selection, I'd like to record a separate episode, because I think that's genuinely exciting to dig into as well. But if you could give a company three or four pointers when they say, okay, I need a multi-cloud or cloud strategy, what would that be?

Marc Achsnich

Get aware of your data. So think: what data do I have? Important to me? And is it something where I can accept issues or not? Other thing: make it a leadership priority. Think: how important is this to me? Something I really want to drive, so I can push it faster? Because in end, status could hit, kill switch could happen, then you've got the problem. So how long do I want to stay viable if something happens? And last, compare whether cost of vendor lock-in or innovations you can generate through it actually stand in proportion to what vendor lock-in cost you.

Joubin Rahimi

I'd say with these three points we're sending you home again. To shed some light on the way the question is framed. Thanks, Marc.

Marc Achsnich

Sure.

Joubin Rahimi

Thanks for watching and listening. And if you have questions or comments, post them below, message us on LinkedIn or via the usual channels on the website. Thanks for watching. Thank you, Marc.

Joubin Rahimi

Thanks, Joubin.