Having your multicloud cake (and eating it too)

Commentary: Sometimes your efforts to reduce costs can end up costing you more in the long run.

“You can’t always get what you want,” the Rolling Stones once sang, but that hasn’t stopped enterprises from trying with multicloud in the hopes that they’ll “get what they need.” That “need” is digital transformation, cost reduction, security and more, but those aspirational goals may be complicated, rather than assisted, by multicloud. As former AWS executive Tim Bray recently wrote, “There are pretty big pay-offs” associated with going all in on a single cloud. Among them? You guessed it: accelerated innovation (i.e., digital transformation), lower costs, improved security, etc.

Is our multicloud reality messing up our multicloud dreams?

We’re all multicloud now

But first, a truism: Every company is multicloud. Sure, a comparative few get there through strategy, but most land there by accident. Companies buy their way into multicloud (M&A). Developers build their way into multicloud (picking best of breed services across clouds rather than imagining one cloud will deliver all the best services). This is the norm, not the exception.

SEE: Research: Managing multicloud in the enterprise; benefits, barriers, and most popular cloud platforms (TechRepublic Premium)

Some organizations spend an inordinate amount of time trying to optimize for “no lock-in.” In so doing, Bray said, they eschew the very benefits that first prompted them to move to the cloud: “You can decide you’re just not up for all these proprietary [cloud service] APIs. But then you’re going to need to hire more people, you won’t be able to deliver solutions as fast, and you’ll (probably) end up spending more money.”

Which is another way of saying that multicloud, however inevitable, may not always be advisable.

A key reason, Bray went on, is that it’s hard to find people proficient in more than one cloud provider. For those that earn this unicorn status of cross-cloud familiarity, there’s a pot of gold waiting, as Google’s Forrest Brazeal has suggested. This multilingual cloud professional, however, remains relatively rare. As such, it’s probably easier to make best use of the cloud talent—already in limited supply—by focusing efforts on one cloud. Or, as Bray put it, “[E]very time you reduce the labor around instance counts and pod sizes and table space and file descriptors and patch levels, you’ve just increased the proportion of your hard-won recruiting wins that go into delivery of business-critical customer-visible features.”

All true. Also true that most organizations are multicloud by accident, and sometimes on purpose.

You keep saying that word …

For those who claim to do so on purpose, why are they doing it? Many turn to multicloud to solve core problems. In a 2022 Hashicorp survey, respondents were asked to pick their primary purposes in going multicloud. The top-four responses?

• Digital Transformation (34%)
• Avoiding single cloud vendor lock-in (30%)
• Cost reduction (28%)
• Scaling (25%)

Ironically, multicloud can complicate these goals, rather than enable them. I wonder if the Hashicorp survey respondents realize this? Because when asked about “primary inhibitors” to cloud success, these same folks cited:

• Cost concerns (51%)
• Security concerns (47%)
• Lack of in-house skills (41%)
• Hybrid cloud complexity (35%)

In other words, the very thing that they’re hoping to reduce costs, for example, may be increasing them. War story: I once worked for a company that made a very public decision to switch from one cloud to another and go “all in” on the second provider. Guess what? It didn’t really work. Instead of decommissioning the old cloud infrastructure, engineers simply spun up new instances on the new provider while never shutting off the old ones (too much risk of crippling existing applications as the new cloud didn’t have like-for-like functionality).

None of which is to say that you shouldn’t prepare for multicloud. You should because the odds approach 100% that you’ll end up there, whether you want to or not. And given that customers are going to use multiple clouds, SaaS vendors and others must of necessity offer multicloud capabilities: You have to enable customers on whichever platform they prefer.

But let’s not pretend that going multicloud is magically going to cut costs, improve resiliency, etc. There is no magic solution to these issues in the private data center, and the public cloud doesn’t offer one, either.

Disclosure: I work for MongoDB, which promotes multicloud as a benefit. My views expressed here are mine alone and don’t reflect those of my employer.