My Case Against the FinOps Phases

????

So, I exchanged a note on this with Phillip Purcell after you posted. Clearly, you wanted to start dialogue on this topic. Mission accomplished! It might be my background in IT service management prior to joining the FinOps ranks, but I don't see any conflict between how the phases are defined today and employing FinOps capabilities effectively. The phases are meant to be iterative and a virtuous cycle of improvement. In ITIL 4, this is the Continual Improvement general management practice. We know that every organization is different, has approached cloud adoption uniquely, have priorities that can't and don't exist in organizations, and may or may not possess components of FinOps capabilities with varying level of maturity. It isn't a one-and-done situation where an organization starts at Inform, move through Optimize, and complete at Operate. Heck, it probably ???????? jump right in for activities under the Operate phase before getting to the Inform or Optimize phases, but I posit that it ??????????'?? ???????????? ????????????. The importance of the three phases is the ??????????????????. Where the rubber meets the road doesn't matter on a wheel - and examining the spokes and how they are ordered misses the point.

The whole "shift left" thing needs to go. This directly requires a left-to-right model, which is far from ideal. Also, if you're not already considering CI/CD as a starting point for all development you're not ready for FinOps. Optimizing implementation and architecture is part and parcel necessary. I agree with you the "phases" is an undesirable term. It implies linear thinking. FinOps is not a process. Finally, the term FinOps itself is nothing but a buzzword now. The FinOps Foundation is doing nothing but elaborating details and additions in the framework for the sake of justifying its existence. It's a bit like the original Mustang auto where Ford kept making it bigger and fatter until it wasn't;t a sports car anymore.

Agree with Stephen Old sentiments. I speak with organizations at various stages of the finops cycle and ideally the finops stages provide a mutually agreed to benchmark of what success looks like. I’d also say that Finops isn’t mutually exclusive from working/assimilating with other organizational disciplines like TBM, DevOps since one of the main goals of Finops is democratizing information (Shift Left) with various stakeholders in an organization to achieve cloud cost efficiency. Getting back to your assessment - as companies mature past the 3 stages I do agree with you that the stages will need to (and are being) updated constantly - I know my guy J.R. Storment is already working on this!

Further to Stephen Old’s point - it’s not about linking Capabilities to Phases. Rather, each phase is addressed while taking the actions related to a Capability. Also, the very nature of a framework is that it is flexible, not rigid as you’d mentioned. It’s guidelines, not a blueprint. Inform - “Gather the data necessary and analyze it.” Optimize - “What can we do based on this data to make things better?” Operate - “Let’s make it happen (or not)” **then return to Inform to confirm results and decide if further actions are beneficial** Now, I’ve absolutely seen marketing across the board that ties capabilities to phases (for example, the legacy Rightsizing Capability often incorrectly gets shoehorned into Optimize). This is an incorrect view as inside the Capability, one still must gather and analyze the data. Also, the Optimize phase isn’t the action phase… that’s Operate. So I’ll say I agree with >50% of what you’re saying here and, in my opinion, we need to hold marketing teams accountable to stop showing this wrong for the sake of simplifying the framework to make it look pretty. Also, FinOps abjectly ISN’T “easy”. Adopting FinOps is all about a change of culture within the organization, which takes time.