Some criticize analyst firm Gartner for being perpetually late in its prognostications, but its declaration that “on-premises is the new legacy” is ahead of the curve. How far ahead of the curve? That remains to be seen, but even as cloud has become the new default for application deployment, Forrester analyst Chris Gardner is correct to argue that “[O]n-premises will be around until the day I retire.” In other words, however much enterprises may want to go “all in” on cloud computing, it’s going to take a long, long time to get there.
The future is now
Gartner’s eulogy for on-premises software stems from its analysis of database revenues over the last seven years. While revenue doesn’t yield perfect insight into the state of the market (much of the open source database market, for example, isn’t counted), it still shows an incredibly pronounced shift from on-premises databases to cloud-based database services. In fact, Gartner continues, if we look at just 2017 through 2018, the two biggest cloud vendors (AWS and Microsoft Azure) accounted for 75% of all database growth in the market.
Wow. Just wow.
The money is in the cloud and, not coincidentally, so is the innovation. Again, Gartner: Innovation “is happening only there [in the cloud], or at the least, in the cloud first.” In fact, “there is an increasing amount of innovation that will never get to on-premises, even when the vendor has on-premises products.”
For these and other reasons, the database market is in serious flux in a way that it hasn’t been for decades. For years Oracle held sway as the undisputed king of database revenue, with Microsoft and IBM following not far behind. AWS now tops IBM and for all its chest-thumping bravado, for the first time Oracle’s perch atop the database market seems shaky. Unfortunately for Oracle and other legacy tech vendors, the more data finds its way to the cloud vendors, the more data gravity will tend to favor these same vendors for applications and other enterprise software.
Of course, we’re nowhere near an all-cloud future. Yes, over time, as Forrester analyst Dave Bartoletti posits, many of those stodgy on-premises workloads will make their way to the cloud. But not now. Not yet. And why? “It’s just good solid stuff that we have to run for a long time but should probably use cloud (in all its forms) to modernize when we finally get the budget and wherewithal to tackle it.”
SEE: Serverless computing: A guide for IT leaders (TechRepublic Premium)
Along the way, the only real revenue growth for on-premises (which marked 32% last year) “is rarely from new on-premises deployments; it is generally due to price increases and forced upgrades undertaken to avoid risk.” So not only does on-premises come with a distinct lack of innovation, but to add insult to that injury, on-premises vendors are squeezing their trapped customers with price hikes.
Even so, there’s ample opportunity for vendors to help enterprises transition to public cloud on their own terms. Additionally, given that over 80% of enterprises espouse (intentionally or not) multicloud environments, those same vendors can play a role in helping enterprises to manage disparate cloud resources. What no vendors can really afford to do is bank on a steady stream of on-premises revenue from databases or other applications.