The Road to FinOpsX 2026

FinOps X 2026 will be held in San Diego next week, and I'm going. I've been around this discipline since before it had a name, and the run-up to the conference motivated me to sit down and dig into how the whole thing came to be.

My inadvertent entry point to FinOps was Adobe in roughly 2012. We had moved our services from the data center to AWS and were quickly among their largest customers, which made us one of the largest customers of a Portland startup called Cloudability. I'll never forget how an errant configuration change on our side broke Cloudability's ingestion of our billing data entirely.

Back then, the whole job of cloud cost management was making sense of an AWS bill. Since then, with the addition of more players in the cloud space, the discipline has had to keep pivoting to serve a steadily broader purpose.

When Nobody Knew What to Call It

The tooling came first. Cloudability launched in 2011 out of Portland, founded by Mat Ellis, JR Storment, and Jon Frisby. CloudHealth Technologies followed in 2012 out of Boston, founded by Joe Kinsella. Both were building products to parse AWS billing data, which at the time was a non-trivial data engineering problem. Apptio, founded in 2007, came at the problem from IT Financial Management and grew into the cloud cost space from there. Pardon my foreshadowing.

Cloudability and CloudHealth were the survivors of a larger cohort. A lot of "cloud-something" companies launched in those years, and the consolidation was inevitable:

• October 2018: VMware acquired CloudHealth for north of $500 million.
• November 2018: Vista Equity took Apptio private for $1.9 billion.
• May 2019: Apptio acquired Cloudability.
• 2023: IBM acquired Apptio.
• January 2026: Flexera acquired both ProsperOps and Chaos Genius, after previously absorbing Spot by NetApp's FinOps capabilities.

The independent FinOps tooling vendor is becoming an endangered species as GitHub and hyperscaler open sourced dashboards clog (and confuse) the space.

The standard story is that the FinOps Foundation defined a discipline. The honest version is that practitioners had already invented solutions to billing problems, vendors built a market around those solutions, and the market is now consolidating around whoever can own the most of the stack.

2019, All at Once

FinOps is a portmanteau of Finance and DevOps (a portmanteau of a portmanteau, as the Foundation's homepage will tell you, and specifically NOT Financial Operations). The cultural model it borrows from is DevOps: cross-functional teams, shared accountability, continuous improvement. JR Storment is generally credited with first presenting the concept publicly at the AWS Public Sector Summit in 2016.

On February 26, 2019, the FinOps Foundation was announced at the Cloud Economic Summit in San Francisco, with eight founding organizations including Atlassian, Autodesk, Spotify, and Nationwide. Storment became Executive Director, and Mike Fuller (Atlassian's longtime FinOps engineer) became CTO.

On December 31, Storment and Fuller published Cloud FinOps with O'Reilly. I still have the marked up copy I used to provide JR feedback. It, alongside my signed Sendmail book, will never be purged from my bookshelf.

By August 2020 the Foundation earned some serious cred and significantly established itself with a move into the Linux Foundation. They were among giants like the CNCF, whose projects range from Kubernetes to Prometheus, and would pre-date their sibling in the PyTorch Foundation. CNCF owns compute, PyTorch owns AI, and FinOps sorts out the bill.

As of mid-2026, the FinOps Foundation is approaching 100,000 individuals and 15,000 companies, including 93 of the Fortune 100.

DBRRT, CUR, and FOCUS

When I was working with Cloudability at Adobe, the format we were all parsing was the DBRRT (Detailed Billing Report with Resources and Tags). It was a CSV. It was enormous. It updated multiple times a day. Which service updated when was a crap-shoot. On Adobe's side, so was tag consistency. Translation: challenges.

AWS replaced DBRRT with CUR (the Cost and Usage Report) in stages through the mid-2010s. CUR was a better data model, but still AWS-specific. Every other cloud provider had its own format, so multi-cloud cost management meant maintaining three or four separate parsing pipelines and hoping none silently broke when a provider quietly changed a column header. Fortunately, I left Adobe as multi-cloud was embraced.

FOCUS (FinOps Open Cost and Usage Specification) is the attempt to solve that at the spec layer. I first heard of it in 2023. It reached version 1.0 GA a year later at FinOps X. AWS, Azure, Google Cloud, and Oracle all launched native FOCUS-formatted billing exports on the same day. That kind of coordinated release across competing cloud providers is rare. Mad props, JR et al! Subsequent releases (1.1, 1.2, 1.3) have expanded scope to SaaS, PaaS, and contract commitment data on a five-to-seven month cadence.

If FOCUS sticks, meaning practitioners actually consume FOCUS-normalized data rather than continuing to hand-roll their own translations, the discipline has a durable cross-vendor standard for the first time in its history. If it doesn't, it becomes another well-intentioned spec that exists more than it gets used. The question of which direction FOCUS goes is open as of 2026. Personally, I'm curious to see how AI and ITAM will come into FOCUS. Thanks, I'm here all week.

FinOps X: Two Notes From Inside

The Foundation's annual conference launched in-person in June 2022 at the JW Marriott in Austin. 454 people showed up. I was one of them. I was scheduled to appear on an AMA panel. It didn't happen, but I was able to connect with a lot of interesting and influential people, a few of which have become lasting friendships.

The 2023 edition moved to San Diego, where it has stayed. Everything felt larger, more polished, less like a practitioner gathering. I was at AWS then, and AWS was not yet a FinOps Foundation sponsor while Google and Microsoft both were. JR made a comment about that at the keynote. It was subtle, but not missed. AWS, the cloud provider whose billing practices had arguably created the entire need for the discipline, was absent from the sponsor list. I did a fair amount of polite head-down hallway navigation at that event.

Attendance ran roughly 1,000 in 2024 and 1,670 in 2025, with 2026 projected north of 2,500. A European edition launched in Barcelona in November 2024 and I'd be lying if I said I was surprised there's been talk about doing it again. After all, it's very Linux Foundation to do US and EA and who doesn't love Barcelona?

Build, Buy, and Shift Left

Adobe eventually internalized its cloud cost management. I may be slightly to blame. At a certain spend level, the math on paying a third-party vendor starts to look worse than the math on hiring a small team to do it in-house. I've always wondered how well that worked out multi-vendor (but didn't dare ask those I left behind).

Build versus buy is one of the hottest topics in FinOps right now, and the consolidation pattern makes the question sharper rather than easier. If you buy a focused tool today, you're buying from a company that may not exist independently in three years. If you build, you carry the maintenance burden forever. The increasingly viable third option is to build a thin in-house layer on top of FOCUS-normalized data and buy point tools for the things that genuinely need specialized infrastructure (commitment optimization, AI cost attribution). That hybrid wasn't really possible in the DBRRT era. FOCUS changes the math.

Running parallel to all of this is what practitioners have been calling shift-left: pushing cost awareness earlier in the engineering cycle so architecture decisions account for spend the same way they account for performance or security. Terraform cost diffs in pull requests, IDE plugins flagging expensive patterns, dashboards built for engineers rather than finance. Whether it sticks depends on whether engineers actually engage with it, which is the same question that decided whether security and observability got real traction in development teams or stayed an afterthought.

The Tension at the Center

As the technical plumbing gets more standardized, the institutional scope is getting more complicated. The State of FinOps reports describe progress, but practitioners who have run a practice know that progress here is asymptotic: you catch the obvious waste, set up the dashboards, negotiate the commitments, and the next quarter brings new waste to catch. Continuous gardening, not a project with a completion date.

The Foundation's expansion into SaaS, licensing, private cloud, and AI is responding to real practitioner pressure. The 2026 State of FinOps survey shows 90 percent of teams managing SaaS, 64 percent managing licensing, 57 percent managing private cloud. The question isn't whether the discipline should address these things; it's whether it can without losing the focus that made it useful in the first place.

AI is the most aggressive driver of that expansion. Two years ago, AI cost management was a concern in 31 percent of FinOps teams; today it's 98. The Foundation has a FinOps for AI certification track, and the tooling vendor market is reorganizing around AI-driven automation. Flexera's January acquisitions of ProsperOps and Chaos Genius were both pitched primarily as agentic FinOps. AI is simultaneously the budget line CFOs need explained and the means vendors are using to automate the work itself.

As if the cloud providers aren't increasing SKU count fast enough.

David Williams' "FinOps is Stuck" Medium piece puts the worry plainly: cloud waste is still rampant, the discipline hasn't solved its founding problem, and broadening the mandate before finishing the original job risks producing a discipline that does many things shallowly rather than one thing deeply. The 2026 framework update moved the Foundation's mission from "managing the Value of Cloud" to "managing the Value of Technology." That expansion could make FinOps the central financial discipline for all infrastructure spending, or it could make it another generic acronym that means different things at different companies.

This was news to me, while doing this research. I'd heard grumblings from time to time, but never outlined like this. The broadening to include SaaS and ITAM does feel a bit like the cloud moving from infrastructure to managed service (and worse). Interesting to think about, but I'm still firmly in the "more good than bad" camp.

Find Me in San Diego

FinOps X 2026 runs June 8-11 in San Diego. Roughly 2,500 practitioners expected, every major cloud provider on stage together, the next FOCUS release on the agenda, and the FinOps for AI certification that was targeted for GA earlier this year.

I'll have my eyes on AI topics most of the time, with the remainder of attention on FOCUS, consolidation, and those hallway conversations you can't help overhearing as you walk between destinations. Of course, I'm looking forward to meeting other practitioners and making new friends, as well.

See you in San Diego, or back here next week.