Don't be a lemming: Cloud-first doesn't mean cloud-only

When 'cloud-first' becomes interpreted as 'cloud-only,' you'll waste a lot of time and money on poorly fitting platforms

“Cloud-first” began as a government concept, one that is still promulgated today. Government CIOs are mandated to look at cloud computing first and foremost when they plan an expansion of their IT footprint. Some have been successful, but for most, it’s been slow going.

The cloud-first strategy has now found its way into Global 5000 companies. Although cost savings is the battle cry, most enterprises try to align their existing and future workloads with mostly public cloud platforms. They are, in essence, taking an approach where the cloud should always be considered. Most of the time, they push away traditional approaches.

But when they hear executives talking cloud-first, most IT leaders will consider it to be a mandate, not a consideration. Indeed, when I hear business execs say “cloud-first,” it typically means “If you don’t go cloud, you have some explaining to do.” That dogmatic interpretation of “cloud-first” as “cloud-only” essentially forces IT to opt for cloud-based platforms, no matter if they are a fit.

Being dogmatically cloud-first is as dangerous as not considering the cloud.

Fit needs to be your priority. I do a great deal of analysis to determine if my clients’ workloads are a fit for the cloud. Although 65 percent of them, on average, may be a fit, the other 35 percent, on average, are not. If you take a dogmatic cloud-first approach, that means you’re going to move about 35 percent of your applications to the wrong platform.

Fit issues come down to a few categories, including security, compliance, governance, performance, and the ability to use platforms or services in the public cloud. Many in IT assume that these services or platforms already exist, but that’s not always true. They end up finding out the hard way that, for example, core networking and database services are missing for the applications they are migrating.

I believe most IT shops follow the correct approach to “cloud-first”: To start, see if the cloud is the best fit and use it; if not, then use an other approach. But I fear that if the concept of cloud-first becomes dogma, organizations will be blind to cloud fit, which is the real measure of how to proceed.

If your company follows the cloud-first dogma blindly, count on wasting a lot of time and money.