While I would argue that cost savings do result from a mature DevOps-driven organization, they certainly won’t come in the short term, nor will they be easy to model up-front. The “savings” are in things like quality of software and better uptime in production. These types of savings are all about “sucking less,” which doesn’t exactly model well in a spreadsheet.
The client I have been working with for a couple of years now, have started "embracing" DevOps about 2-3 years ago. Yet, they're still not able to scale, have a painfully slow (and sometimes manual) release process, with all the paper trail and auditing of good old waterfall times, policies which stem from times when they ran their own data centres, and most projects are already so far over budget that getting a security patch approved takes longer than it ever took before.
They've burned a lot of money and aren't anywhere closer to understanding the concept of DevOps. They fell for "cloud is cheaper, everything is on demand, and releases and testing will be fully automated", completely oblivious to any advice suggesting that none of that is going to come cheap. The biggest cost in DevOps is the shift in thinking. Release and approval processes have to be changed, the way people develop need to change, most applications need to be changed to cope with the dynamic nature of cloud and devops. If there's only one cog in the machine that doesn't spin along nicely, the whole thing grinds to a halt and burns money, which is exactly what's been happening there.
It's a huge and very old organisation. It won't put them out of business, but eventually somebody will pull the plug and say: right, doesn't work for us. Stop burning money or else! The result might just be rowing back to a similar setup as they had before. Interesting to watch, though putting out fires while being hamstrung all the time does get a little bit annoying.
The development and devops teams (yes, they are somehow different teams) have been significantly increased over the last 18 months, although the number of projects decreased... Yet, little progress. You could argue they are sucking more than they did before. Also difficult to explain to a spreadsheet! ;-)
They are squarely in the 90% unless a miracle happens. Given the size of an organisation like that, realisation sets in at a much slower rate. But 2018 sounds like an achievable goal to bin the idea of devops in that company.
Anonymous for obvious reasons.