Finops can show where all your cloud spending goes, but sometimes penny-wise is pound-foolish. Account for human costs and monitor all your clouds.
These days cloud computing costs typically run about 20% of total IT spending. An enterprise only needs one outrageous cloud bill to wake up and smell the waste. That’s why cloud spending now receives much more scrutiny, and enterprises demand more discipline with cloud costs. Finops provides the ability to monitor and optimize cloud spending. Hence, it’s becoming a large part of any cloud deployment.
Finops is a good thing, but not all cloud finops programs are the same. Let’s talk about the most common missing or misunderstood finops features:
Failure to understand the business value derived from cloud spending. Many cloud finops programs and their users see any type of savings as a good thing since it drives a better bottom line.
The trouble comes when they don’t consider business value generation. Some cloud savings may inadvertently reduce or eliminate an important soft business value. For example, finops might recommend limiting the use of cloud-based AI systems due to higher costs without understanding that those key systems can realize a 100-fold return from any AI spending. When reviewed, that $0.10 the finops team saved actually cost $10.00 in unrealized business value.
Of course, unrealized business value metrics are often the most difficult to define and track. Finops programs and teams need more than a rudimentary understanding of cloud spending and how to reduce that spending, but they also need to understand the ties between business value and specific types of spending.
Failure to consider human costs. The costs of humans also need to be factored into cloud spending. Many times, they are not.
This will get you in trouble if cuts to cloud spending require more human hours to reach the same net effect. Thus, a net negative benefit. The finops team can’t understand this unless they look at the number of human hours spent on the same business processes before and after savings adjustments. Hopefully, finops runs virtual “what if” scenarios before implementation. You need to monitor and optimize both to return the most value to the business.
Failure to monitor all public cloud providers the enterprise uses. It confuses me when an enterprise has two or three different public cloud providers but only monitors the costs of a single provider.
This is a remnant from the single provider days. Most enterprises started out on a single cloud provider and then built a finops program around that provider. Many even standardized on finops tools that are proprietary to that provider and typically lack the ability to monitor or analyze spending on other cloud providers when they arrive on the scene—and those additional clouds always show up. Those who build systems within enterprises need the ability to pursue best-of-breed solutions supplied by other providers.
The lesson here is that you need to monitor and govern spending across all the different cloud providers, even before those providers become part of IT’s responsibility. Monitoring a single cloud provider, even if that provider has 80% of your cloud services, means that you are only getting part of the story. That story is almost guaranteed to turn frightening when your enterprise moves to multicloud.
We’re just getting started with cloud finops programs, although cloud has been around for a long time. Like many parts of the cloud, do finops right the first time, or expect some rather costly mistakes.
Courtesy: https://www.infoworld.com/
No comments:
Post a Comment