I just had a customer ask, “Is Google Cloud Platform (GCP) cheaper than my on-premise setup?” I knew that the answer was an unequivocal “yes, of course”.
GCP charges you only for what you use, is spending $10b+ a year on infrastructure (are you getting a better deal than Google procurement?), operates at extraordinarily high utilization, and is competing to win market-share. Think Magic 8ball: signs point to yes!
But this is JOCKLEY! Proof is the standard we use.
Unfortunately, Google’s sales enablement, marketing and analyst relations have not yet sponsored direct analysis as to whether GCP is cheaper than an on-premise setup. (Hey Google Cloud homies! If you want me to pull up a team to do one, just let me know!)
For a refresher, check out this wikipedia article.
Let’s do that fun story problem stuff you did in grade school, and replace those variable names for the solutions we’re talking about:
Conveniently, my friends at AWS (howdy peculiar folk!) have absolutely sponsored this kind of detailed TCO analysis, working with the outstanding folks at IDC.
Their report says:
“The five-year total cost of ownership (TCO) of developing, deploying, and managing critical applications in Amazon cloud infrastructure represents a 70% savings compared with deploying the same resources on-premise or in hosted environments. The findings showed a 626% ROI over five years.”
See now that’s what I like, some real numbers to work with. Thanks IDC!
There are lots of analyses (this one, or that, or hey this, or perhaps this other one) that show GCP is cheaper than AWS, but we’re into the formal, structured methodology, rigorous modelling around here. Which of course reminds me of a paper I worked on together with fine folks at ESG, which says a similar thing:
“ESG Lab validated that for a mature application in production, GCP’s realistic, cumulative three-year cost was 35%-53% lower than the various AWS pricing options.”
So let’s recap:
IDC says you save 70% by moving from on-premise to AWS.
ESG says you save at least 35% moving from AWS to GCP, in the worst case.
I bet you can already smell the formula that’s brewing
100% x (1-.70) x (1-.35) = 19.5%.
This means that GCP will be on average, just slightly less than one-fifth of the cost of on prem. Transitive property to the rescue!
Google Cloud Platform is going to be cheaper than your data center.
As always, the specifics matter; no two environments are the same, and perhaps you got better deals on equipment or software or services than either of these analyses assume. Could be you’ll only save ½, and oh my what a bummer that would be. Maybe you got a worse than average deal, or your workloads are a better than average fit for cloud: it could be even better!
Maybe it’s worth taking some time with us to do a more detailed analysis on what we think your specific results would be in a move to the cloud. We’re pretty good at helping customers figure this kind of move out; it’s all we do!
In fact, I’ll make you a deal: Assuming normal caveats, we’ll do that analysis for you for free. Just reach out!