There is well referenced cloud ROI article/survey that was done by ‘Information Week’ a couple of years ago and, among other things, the article compares and contrasts Airbnb and General Electric. Most of us are firmly in between the Airbnb ROI perspective (what a waste of time, all cloud because I have no legacy, and who the heck knows which direction we will need to pivot next year) and the GE ROI perspective (we measure everything and I mean everything, with 34 data centers and 9000 applications this could take a while). So how important and useful is doing a Cloud ROI?
Well like the consultant once (always?) said, it depends. If the end game of the ROI is get to a thumbs up/thumbs down decision, then probably a waste of time. I don’t think I am going out on a limb here but, shhhh, the Cloud is not a fad. If the end game of the ROI is really developing the business case with specific benefits, costs, and risks, well then you have the beginning of a migration road map and a measuring stick. Well done.
I like to define the Cloud = Infrastructure + Apps + Data + Risk (Reduction).
So let’s begin with infrastructure where many companies start their cloud migration. First things first: don’t open your wallet! The next time you are faced with a big infrastructure upgrade is a good time to think Cloud. There is a lot of talk about moving from CapEx to OpEx – moving from buying capital and depreciating it versus just paying for usage as you go. (Side note: this is kind of funny since I remember not too many years ago when everyone was looking to capitalize as much as possible in IT). But the real ROI is realizing your current on premise infrastructure is by definition a big step cost curve where you are either operating well over required capacity or operating well under required capacity. Either way not good. The cloud allows you to much more closely match your infrastructure cost curve to your systems usage curve.
Next Applications. There are differences depending on whether they are built or bought but same general principle. Lifting and shifting an application to the cloud as-is (Infrastructure as a Service) can provide a modest return fairly quickly. But a major refactoring (by you or your provider – going to either Platform as a Service or Software as a Service), may provide a greater RIO albeit with a longer duration to pay back. For the latter, the application really needs to be re-thought both technically the architecture and functionally taking advantage of available cloud services.
Data. This is an area likely less about tangible costs/benefits and more about softer benefits. How much is better insight worth to your business? Before the cloud, Watson or Cortana or whomever, would have been the province of only the biggest that could afford this kind of stuff. But now you can take your CRM data already in the cloud, point that data set at the machine learning service and within days and very little money (cause you’re renting it right?), start getting some insight on the next best cross sell.
Risk encompasses security, downtime, reputation – all those things that can lead to business disruptions both directly (I can’t get to your site) and indirectly (I don’t want to use your site). Reducing risk is certainly worth something in your ROI – people pay for insurance after all. Of course opinion can be tainted with a few well publicized Cloud outages but consider that in 2016 the Microsoft Azure Cloud (for example) had 270 minutes of down time. Note that is across 30+ regions so if you (or your provider) has architected your app to be fault tolerant across regions you are now looking at something like 99.9979% uptime.
So if you are going to do an ROI, use it to ensure you are not missing the low hanging fruit. And that alone could fund the bigger opportunity to be faster, more agile, and more innovative and ultimately help level the playing field in your favor.