A move to the cloud is no simple calculation

Services like email are easy to justify moving to the cloud -- then there's the rest of your infrastructure

It's easy to assess the pitch for moving services like email from on-premises deployment to the cloud. If the effort saved is sufficient, the cost is appropriate, and the risk is acceptable, you do it. Many companies have already made that decision for the likes of Exchange, retiring lots of servers in the process.

But what about the remaining infrastructure? Is it time to look at moving the rest of it to the cloud?

As the economy continues to improve after the 2006-2009 recession and the halting progress in the years following, you likely have a lot of aging infrastructure. You can spend money on replacing it in your data center or move it elsewhere, such as to a public cloud or to a virtualized environment in a vendor's data center.

When you deploy your internal technology, you make the capital investment up front. Hardware vendors love to show you that doing so saves you money over a few years versus renting that infrastructure by sending it off to a vendor data center or cloud. Cloud and service providers will argue that by not buying your own, you'll save more on operating expenses over the long term; plus, you don't have to get all that capital upfront (or finance it).

If cost savings is your only motivation for moving to the cloud, you'll be disappointed. There are hidden costs to migrating away from your data center, but you won't hear in providers' sales pitches. Also, the learning curve exacts its own price. (For smaller businesses, cloud pricing can be too high given the minimums often required.) No cloud vendor's ROI calculations address those realities.

Instead, look at the value the cloud can provide in letting your staff focus on what matters most to your business's success.

You should look at the flexibility of being able to increase capacity on the fly. By using the cloud, you won't have to forecast exactly how much capacity you'll consume. The cloud's pay-as-you-grow (or pay-as-you-need) model tends to work better in burst situations, when you could use that extra capacity for a brief time during the day, week, month, or year.

There is also the value of having bottomless storage that's typically cheaper than on-premises drives -- and you can release the storage at will if you need less of it in the cloud.

Are there risks to moving your infrastructure to the cloud? Certainly. A backhoe could cut your Internet connection, and you're now disconnected from your data and computing. All cloud vendors have partial or full outages (granted, so do your own data centers). When there is an outage or a security breach, you have no way to intervene in the process as you would on-premises.

And if you have a well-trained, super IT team now handling your servers and your company's needs, a move to a faceless IT team in the cloud may be a step down.

That's why your decision must be based on individual requirements and realities. Internally, that includes the existing hardware, the quality of the existing IT team, the effort required to maintain the existing environment, the current and desired vendors. All of this must be determined before you consider the next step. On the cloud end, it includes the vendor mix, the connectivity reliability and throughput, redundancy and recovery options, the technology stack, and the quality of the virtual IT team.

Eventually, everyone will simply plug in and IT will be a utility. An on-premises server will look like a jukebox does in a diner. Until that day, there's a legitimate debate and assessment to be had.


Copyright © 2015 IDG Communications, Inc.

How to choose a low-code development platform