During the last Re:Invent, AWS repeated it’s mantra: multi-cloud doesn’t really exist.I have to agree it’s a great mantra. Unfortunately, then comes life. Amazon is using Azure. Google is using AWS. These are facts. Are these companies short of engineering talent or cash, or animosity toward the competition? No, no and hell no! They simply deal with the reality that was forced on them. In this post, I’ll review the reasons that may force your organization to face this reality.
Amazon has become a customer of Azure when they acquired Whole foods. Google of AWS when they acquired Waze. If you think this problem is not related to your organization, think again. The number and size of M&As is a growing trend, and the probability you will find yourself on one end of the acquisition table is higher than ever. I won’t review the many variables that causes the M&A trend here, but it’s not going away soon.
As M&As are mostly executed in search of synergies, and IT is one of the key areas these synergies can be attained, one should assume there is a strong motivation to consolidate the IT of the merged companies. While AWS leads the market, Azure and Google are growing faster, especially in some areas. Assuming the market is divided between these three at 50%, 30% and 20%, market share (ignoring Oracle, IBM and the rest), the chances are in favor of any merged organization ending up in a multi cloud reality. Whether this is a wanted state or not is beside the point, and CIOs and CTOs of the merged organizations will need to deal with it.
Location and regulation constraints
Generally speaking, you can ignore the fact that most cloud service are contained in a specific region/location. However, if the application or data is regulated, e.g. GDPR, you’ll find yourself facing a new set of challenges relating to application span, data silos and an overall competing need to segregate and standardize environments. The cloud vendors offer many tools that help with these problems, but not a solution.
Moreover, when expanding to a new region, you might find yourself in a new environment that is lacking some of the functionality that your original location had. Add to that the fact that not every cloud provider has a datacenter in every country, and you might find yourself forced to use other cloud providers or, heavens forbid, your own servers in colocation datacenters. Whatever the situation, the simple, no brainer answers simply don’t apply.
Shadow IT causes unplanned cloud adoption of all sorts
One of the key strengths of the cloud is how available it is. Provide your credit card details and consume. If your organization is already using the cloud, it doesn’t make sense for a business unit to use other cloud vendors, or does it?
The most trivial reason for using other cloud vendors is the knowledge of the dev team. Organizations don’t want to learn multiple APIs and guess what, developers don’t want to either. If a developer is used to an API, has code examples for the API and development best practices around that API (“is this transaction guaranteed or eventually consistent for my client or any of the consuming streams?”), they would not be eager to learn a new way of doing things, especially when TTM is key and everybody is rightfully pushing to an agile “just do it” way of thinking. As this “competing” knowledge is gained by the team when acquiring new talent, you can be assured your team’s knowledge will always be somewhat “tainted”.
Survival of the fittest cloud services
AWS claims nobody can compete with their huge variety of services and that’s why customers will use AWS cloud as a single cloud provider. If only life was that simple. Let’s first address this claim from a pure product appeal perspective: nobody cares how many features you have, but rather look at what value your features bring to the table. I would go as far as saying that with the noise levels today, customers want less services and features. They want more solutions to their problems. With an announcement every 3 days, can AWS really claim it creates value to customers in a unique manner? In my humble opinion, AWS Lambda was a real value creator which is why the competing vendors closed the gap fast, and nowadays, a similar feature is offered by every cloud provider.
Does this mean that all cloud vendors will be the same? Quite the opposite. This will force the cloud vendors into a strategic war over specific niche markets with specific solutions tailored to specific needs. The factors setting their course would be their ability to acquire talent and customer traction in those specific fields. This would bring in specialization areas, which in turn will create even more multi-cloud adoption, not because people want it, but because they’ll be forced to use best of breed since the differences between specific cloud services will too big to ignore.
Ok. That’s life. Now what?
If you’re still reading this, you are at risk of deteriorating to existential questions and that is not my goal. Quite the opposite. I claim that since multi-cloud is a reality, it requires a mitigation plan. CIOs and CTOs should have a good sense of how to deal with it, whether they’ve already faced it in their organization or not. The ones who have a good cloud strategy, together with solid multi-cloud mitigation plans will be able to act in a more flexible and agile manner, and will increase their organization’s chances of winning.