The FinOps Framework 2026: From Cloud Cost to Technology Value
For most of its life, FinOps meant cloud financial operations. The 2026 edition of the FinOps Framework, and the announcements at FinOps X in June, finish a shift that's been coming for two years. FinOps is now about the business value of all technology spend: cloud, AI, SaaS, data platforms and the data centre. Each of these bills in a different way, and each has a different owner. This piece covers what changed, and how to turn it into one return-on-investment story that every persona can contribute to.
What changed in the Framework (March 2026)
The FinOps Foundation's 2026 Framework update, published 19 March 2026, makes four changes worth knowing:
- A new capability: Executive Strategy Alignment. It sits in the "Manage the FinOps Practice" domain and connects technology value to business strategy, so that leaders can compare options and govern investment.
- "Cloud" removed from capability names. For example, Workload Optimization became Usage Optimization, Architecting for Cloud became Architecting & Workload Placement, and Cloud Sustainability became Sustainability.
- Scopes deepened. A Scope is a defined segment of spending across technology categories, aligned to a business construct such as a product, cost centre or environment. The update adds guidance on how Scopes are started, run and closed.
- Technology Category pages for Public Cloud, SaaS, Data Center, Data Cloud Platforms and AI, plus named converging disciplines: IT asset management, IT financial management, IT service management, sustainability, security and enterprise architecture.
The definition of FinOps itself now speaks of maximising "the business value of technology", not of cloud.
What was announced at FinOps X 2026 (June)
- Token economics took centre stage. The Day 1 keynote (9 June 2026) framed it as turning energy and capital into AI tokens, then consuming them efficiently for value. The FinOps Foundation and the Linux Foundation announced their intent to form a Tokenomics Foundation to create open standards for AI billing.
- FOCUS 1.4 (ratified 4 June 2026) added Invoice Detail and Billing Period datasets, so usage can be reconciled to the actual invoice. It also expanded commitment data, with a column showing which commitment program each charge could qualify for.
- Agentic FinOps. The Day 2 keynote presented a Crawl, Walk, Run model for moving from automated alerts to agentic cost management, with vendors announcing agents and AI cost features alongside it.
Same framework, five different billing models
The practical challenge is that each technology category bills on a different model, has a different natural owner, and pulls a different ROI lever. Treating them all like the public cloud is how AI and SaaS spend escapes scrutiny.
| Category | How it bills | Who usually owns it | The main ROI lever |
|---|---|---|---|
| Public cloud | Resource-time (instance-hours, GB-months), plus commitments | Engineering, with FinOps and procurement | Remove waste, rightsize, commit the floor (2026 commitment changes) |
| AI | Tokens (input, output, cached) or provisioned throughput | Product and data science | Cost per successful output; cap output, cache prompts, right-size models (FinOps for AI) |
| SaaS | Seats and licences, often annual | Business owners and IT, through procurement | Reclaim unused seats, right-tier licences before renewal |
| Data cloud platforms | Consumption credits (compute and storage) | Data engineering | Warehouse sizing, auto-suspend, query efficiency |
| Data centre | Capital spend, depreciation, power and space | Infrastructure and finance | Utilisation of owned capacity; workload placement decisions |
The ROI chain: one number per persona, one story per Scope
Executive Strategy Alignment is the Framework's way of saying that the ROI conversation now belongs at leadership level. Leaders can only have that conversation if every persona feeds it one honest number. Here's the chain, from the work at the bottom to the decision at the top:
| Persona | The question they answer | Their number | Where it comes from |
|---|---|---|---|
| Engineering | Are we using what we pay for? | Waste as % of spend; rightsizing realised | Usage Optimization; assessments; provider recommendations |
| FinOps practitioner | Can we see and allocate all of it? | % of spend allocated to a Scope; commitment coverage | Tagging, allocation rules, FOCUS data |
| Procurement | Are we paying the best rate? | Effective savings rate; renewals in the next 6 months | Commitment and contract data (FOCUS Contract Commitment) |
| Finance | Does the bill match what we forecast? | Forecast accuracy; invoice reconciled to usage | Budgets, forecasts, FOCUS 1.4 invoice data |
| Product | What does it cost to serve a customer or an outcome? | Unit cost (per customer, transaction, successful AI output) | Allocated cost ÷ business metric (unit economics) |
| Leadership | Are we investing in the right things? | Value per Scope against total technology spend, and its trend | The numbers above, rolled up per Scope |
The key move is measuring per Scope, not per provider. A product's true cost is its share of cloud compute plus its AI tokens plus its SaaS seats plus its data-platform credits. Report only the cloud line and an AI-heavy product looks cheaper than it is, which is exactly the blind spot the 2026 Framework is designed to close.
A worked example. A customer-support product (one Scope) runs on $18,000 a month of cloud, $9,000 of AI tokens and $3,000 of SaaS seats, so $30,000 in total. It resolves 60,000 tickets a month, which works out at $0.50 per resolved ticket. Engineering removes $2,000 of idle compute, product caps output tokens and saves $3,000, and procurement trims $500 of unused seats. The new cost is $24,500, or about $0.41 per ticket, an 18% improvement, and leadership sees it as one number against one product. Illustrative figures.
How to start (Crawl, not Run)
- Pick three Scopes that matter to the business: your biggest product, your fastest-growing AI feature, and one shared platform.
- Pull all technology spend into each Scope, not just cloud. Start rough. An approximate AI and SaaS share beats leaving them out.
- Agree one business metric per Scope (customers, transactions, successful outputs) and publish the unit cost monthly.
- Give each persona their one number from the table above, and an owner.
- Only then automate. Agentic tools are most useful once the Scopes and metrics exist for them to act on.
Where are you today? Try the free FinOps maturity self-assessment (Crawl, Walk, Run), or run the CloudFinOpsKit assessment for a priced view of waste, commitments, allocation and AI spend across Azure, AWS and Google Cloud, from your actual bill.
What this article doesn't cover
The full Framework (domains, every capability and maturity criteria) is on finops.org, and certifications are covered there too. The Tokenomics Foundation was announced as an intent, so its remit may change. CloudFinOpsKit measures cloud spend and AI billed through Azure, AWS and Google Cloud. It doesn't measure SaaS or data-centre costs, which you'll need to bring in from their own sources.
FAQ
What changed in the FinOps Framework in 2026?
The FinOps Foundation's 2026 update (19 March 2026) added a new capability, Executive Strategy Alignment; renamed several capabilities to drop cloud-specific wording (for example Workload Optimization became Usage Optimization); deepened the Scopes construct; and added Technology Category pages for Public Cloud, SaaS, Data Center, Data Cloud Platforms and AI.
What is a FinOps Scope?
A defined segment of technology spending, across categories such as public cloud, SaaS and AI, aligned to a business construct like a product, a cost centre or an environment. It's the unit you apply FinOps to, and the natural unit for measuring return on investment.
What is the Tokenomics Foundation?
At FinOps X in June 2026, the FinOps Foundation and the Linux Foundation announced their intent to form a Tokenomics Foundation, a neutral place for practitioners and suppliers to create open standards for AI billing. It was announced as an intent, so its final shape may change.
How do you measure FinOps ROI?
Per Scope, not per cloud. For each product or cost centre, track the value it produces (revenue, customers served, successful AI outputs) against all the technology spend that supports it, across cloud, AI, SaaS and data platforms. Then report unit cost and its trend, and the savings realised against those identified.
Sources: FinOps Framework 2026 · FinOps X 2026 Day 1 · FinOps X 2026 Day 2 · FOCUS 1.4
Related reading: cloud commitments in 2026 · FinOps personas and how they work together · FinOps KPIs · cost per successful output