Is on-premises being killed by the cloud? Not necessarily.

On first view, the biggest story for public-sector IT in recent years is the cloud. Look at some of the headline statistics, and the conclusion seems obvious: on-premises is on its way out, and the cloud is the future. Some of the numbers are striking:

• Eighty-two percent of public-sector cloud adopters will increase spending on cloud computing this year, according to a recent survey.
• Total spending on IT infrastructure projects in 2016 that were deployed in cloud environments grew by almost 20 percent, according to IDC.
• The same study found that spending on traditional on-premises technology declined by 4 percent.
• By 2018, global public IT cloud service revenue is predicted to reach $127 billion.
• There’s been a huge shift in preferences from IT buyers towards the cloud in recent years, according to Software Advice. In 2008, 88 percent of buyers (who said they had a preference) preferred on-premises solutions, but by 2014, 87 percent were now saying they preferred cloud solutions.

There are plenty of statistics that bear out this conclusion, as well as firsthand experience. At a trade show in San Francisco, I was speaking to a specialist who told me he almost always recommends the cloud as a migration destination for new customers. So has on-premises technology been killed by the cloud?

Well, not necessarily.

A more nuanced picture

There’s an imperative for cloud providers to tout the popularity of their products. That’s what good marketing is all about. Bearing that in mind, is the cloud quite as popular as the leading providers would have us believe? I’m not so sure.

Take IDC’s research from above. Yes, the data shows that public cloud investments are growing, but it also predicts that even by 2020, more than 50 percent of spending on IT infrastructure will remain on traditional data centers. And the same goes for Software Advice’s top finding -- while the number of buyers saying that the cloud is their preference is on the rise, the vast majority actually say they have no preference between the cloud and on-premises.

Certainly the cloud is growing, yet in reality things are much more nuanced. If we look at the headline figures, it’s possible to see the cloud as perhaps more significant than it is. This is problematic because it can lead people to rush decisions that aren’t necessarily right for their organization, just because they think they must keep up with the latest trends. It’s also worth simply asking who benefits by the hype about the cloud.

Cloud clearly is a great innovation and offers countless benefits to government agencies, but we should avoid drawing conclusions about the death of on-premises. Below are three key reasons why I think that on-premises computing hasn’t run its course yet.

The political aspect

The term ‘cloud’ is deceptive -- it makes it seem like the servers and the information stored in them are not to be found in a specific location. In reality they are, of course, which raises questions about localization of data. Certain countries (Germany is a standout example) have extremely strict laws about what federal agencies can do with data and where it can be stored. As a result, many German organizations (both public and private) simply cannot store any data in cloud environments outside of national boundaries. Similarly, in countries that are not so friendly with the United States or the European Union, for example, there is often anxiety about storing data in cloud data centers -- the majority of which are to be found in Europe and North America (according to this map).

Who’s buying?

Another nuance regarding the future of on-premises is to ask who is actually making the buying decisions. For instance, if chief financial officers have the final decision, I would guess that they will more frequently opt for the cloud. It is cheaper, easier to set up and means they don’t have to pay so many people to monitor and maintain servers.

However, if the primary buyer is the chief technology officer, the preference will generally be for on-premises for as long as possible because CTOs tend to have more concerns about storing agency data in a public cloud that they don’t control.

That’s not to say all CTOs are against the cloud, of course. Many senior IT staff I speak to would love to offload the responsibility for maintaining servers in their office. If there’s some physical breakdown with an internal server that hasn’t been properly managed, the time and energy spent resolving the issue can be frustrating. And this is certainly part of the reason that private cloud providers are a popular alternative. An organization can pay for rack space in a local, managed data center and still have full control of what happens on the server, without concerns related to storing that data in a huge public cloud.

The point is this: Not everyone is a convert to the cloud just yet, and some may never be -- meaning on-premises still has many years left ahead of it. Others will opt for a mix of on-premises, private cloud and some public cloud --it all depends on the people who make decisions in the organization and what their priorities are.

Agency size

A final nuance here is the size of the organization. It’s one thing for a small, local government agency (or even a new team within a larger organization) to opt for cloud-based operations. It can save on the initial hardware costs, provide large amounts of cheap storage right away, and it’s just easy to set up. However, the picture just isn’t the same at, say, a huge institution like the tax office. An on-premises platform that’s  generally working well provides  little motivation to migrate everything to Office 365 or anywhere else.

Because here’s the thing that we don’t hear about so often: An on-premises-to-cloud migration can be an enormous project. It’s one thing to migrate the content of staff inboxes to Exchange online, it’s a another thing entirely to migrate applications and other business systems. For bigger government agencies, this can easily amount to a two-year long project that features fees for technology, consultancy and change management.

Many CTOs and CIOs feel uncomfortable making this leap just yet. If everything is working now, it can feel like a shot in the dark to move. What’s more, we’ve seen how much the cloud has evolved in the last couple of years. More risk-averse decision makers may want to wait and see if the technology improves before committing to the cloud -- or wait for a better compromise solution to emerge.

A multi-layered compromise

When the cloud first emerged, it was widely touted as the next big thing that would change the way all organizations work. A few years later, however, and that change hasn’t exactly happened. Many customers simply weren’t sold on the cloud, and so now companies like Microsoft have adapted and are providing hybrid-ready on-premises platforms like SharePoint 2016.

For me, this is the kind of compromise I think we will see more often --  where organizations build out a mix of legacy on-premises, private cloud and various public cloud environments. And why not? Perhaps we’ll even see some next-generation on-premises tools in the coming years too.