FinOps
For the last 3 years running, Flexera’s State of the Cloud report has listed Managing Cloud Spend as the top concern from companies using the cloud. Similar results can be found in the FinOps Foundation’s annual State of FinOps survey. Telling, but not entirely surprising. Many organizations struggle with organic growth of their cloud usage and the switch from what was typically a Capital Expenditure (CapEx) to one that is purely an Operational Expense (OpEx).
This has led to the growth of a new set of principles and practices labeled FinOps. A portmanteau of Finance and DevOps, which is a portmanteau of Development and Operations (this is getting to be quite the party). While there are other dimensions that we will explore, at its core FinOps is about transparency, attribution of the costs of using The Cloud, and confidence when making decisions that impact cloud spend.
That is why the FinOps Foundation was founded in 2019. They define FinOps as “an evolving cloud financial management discipline and cultural practice that enables organizations to get maximum business value by helping engineering, finance, technology, and business teams to collaborate on data-driven spending decisions.” Quite a mouthful.
The FinOps Foundation promotes a Crawl, Walk, Run approach to embracing and implementing FinOps within an organization. We prefer a slightly different mantra, which we think is more accurate: Inform, Optimize, Operate.
Inform
FinOps is not about making the cloud cheaper. Moving to the cloud will not be cheaper. At least not initially. If that is what McKinsey or some other SI tells you, eat the free lunch, say thank you, and then throw their deck away. That doesn’t mean that we promote a wild wild west approach. The key first step to control costs is for everyone to see what the actual costs are.
If you aren’t already familiar with guardrails and tagging (and you should be), then FinOps can be your introduction to those very useful concepts. All the major cloud providers allow you to create policies that are applied to the services that you are consuming. Simple things like requiring a project/program name/number so that you can track the spend. Limiting the size of the compute that you purchase (do you really need that super-sized?). Automatically turning things off at “night” and the weekends. Lifecycling policies that delete old snapshots. We could (and will in later blogs) go on for hours about the importance of policy to cloud usage and security.
Even without tagging, all providers provide dashboards and visualizations that allow you to view your spend. They are typically at the account level. If you have multiple accounts (and you probably do), this is also the time to understand your cloud provider’s master account/billing strategy. With tagging, you can now start to allocate those costs appropriately.
Optimize
Okay, you now know what you are spending and on what. In parallel, you should also have started a collaboration between the Finance, Business, Development, and Operations teams. Every group, whether you organize by platform, product, system, or whatever needs to understand and take responsibility for their own cloud spend. What is the business value that is being derived from the spend? How do you give your Exec, Finance, and Engineering teams confidence that they’re doing the right thing for the business?
Part of the paradigm shift is that, done correctly, you will be in a position to fairly precisely calculate the ROI of your systems running in the cloud. In the old days of on-prem data centers, everyone wanted to do it, and there was no shortage of methodologies, but they all failed. Even the most complex methodologies couldn’t resolve the issue of how to allocate all the “shared” costs of running a data center down to business lines, let alone individual products.
Using FinOps practices and real data from your providers, you can accurately allocate the costs of the infrastructure and services down to specific products. You can also leverage centralized economies-of-scale and commitment-based discounts to significantly reduce spend. This is where the actual “wow” moment of FinOps happens. Infrastructure costs stop being just “cost” and become part of the product ROI, much like software development spend that went into developing this product. This has two significant implications:
It gives an organization a better picture of the product ROI, bringing greater clarity to investment decisions.
It gives a product team a much better incentive to manage the costs and make decisions driven by an “economic view”.
But before you get there, you need to be informed first. It is much easier when things are done right from the outset of the cloud adoption - proper tagging is implemented, guardrails are in place, the FinOps team is there, and the tool exists to produce reporting. That, unfortunately, has rarely been the case and a lot of companies are now playing catch-up. Do not despair - better late than never. It is still possible to introduce FinOps. Just expect to take care of some basics first, before you can achieve the Operate phase.
Operate
Proper tagging, guardrails, and cloud policies are not just useful for understanding your cloud spend, but they begin to enable a plethora of advantages from added security, to other cross-cutting concerns, to consistency, to user control. But more on that later…