The Cloud FinOps Foundation defines three stages of FinOps maturity: Crawl, Walk, and Run. The framework is useful, but the transitions between stages are harder than the labels suggest — and most organizations that think they’re Walking are still Crawling in the ways that matter most to finance.
This is not a criticism. FinOps maturity develops unevenly. An organization can have excellent tagging discipline (Walk-level) while still receiving monthly cloud bills with no cost attribution (Crawl-level). The diagnostic question isn’t “what stage are we?” but “where specifically are the gaps, and what do they cost us?”
The Crawl stage: getting visibility
The Crawl stage is defined by the absence of cost visibility. Not the absence of data — AWS provides extensive cost data — but the absence of processes and tools to make that data actionable.
Signs you are in the Crawl stage:
- You learn about cost overruns when the monthly bill arrives, not before
- You cannot answer “which team or product generated this cost” without significant manual work
- Cloud costs appear in finance as a single line item with no breakdown
- Tagging is inconsistent and unenforced
- There is no designated person or team responsible for cloud cost management
The primary output of the Crawl stage is cost visibility: understanding what you’re spending, on what, and by whom. Without this foundation, every subsequent improvement is based on incomplete information.
The Crawl-to-Walk transition is blocked in most organizations by the absence of an allocation methodology. You can buy tools and hire people, but until you have a tagging taxonomy, enforcement mechanisms, and a process for attributing shared costs, you cannot achieve team-level cost accountability.
The Walk stage: accountability
The Walk stage is characterized by cost accountability at the team or product level. Engineers know what their infrastructure costs. Budget owners receive cost reports for their area and are expected to explain variances. Finance and engineering share a cost model that both sides believe is accurate.
Signs you are in the Walk stage:
- Cloud costs are attributed to teams with sufficient accuracy for budget management (85%+)
- Anomaly detection is active and alerts reach the right owner within hours
- Reserved Instance and Savings Plan coverage is tracked and actively managed
- There is a FinOps function — either a dedicated team or a named owner with finance involvement
- Optimization opportunities are identified, tracked, and implemented at a meaningful rate
The Walk stage is where most FinOps programs stall. The common failure mode: the attribution and alerting machinery is built, but the savings don’t materialize. Findings are identified and not implemented. Optimization recommendations are acknowledged and not acted on.
The Walk-to-Run transition requires closing the accountability loop. Identifying a savings opportunity is not the same as realizing it. Finance teams that operate at the Walk level can tell you that $200,000 in annual savings was identified last quarter. Run-level FinOps programs can tell you that $143,000 was actually realized, with evidence.
The Run stage: optimization as a practice
The Run stage is characterized by optimization that’s continuous, measurable, and verifiable. It’s not a project; it’s a capability.
Signs you are in the Run stage:
- Cloud cost forecasting is accurate to within 5% for monthly spend
- Unit economics (cost per customer, cost per transaction) are tracked and improving
- Savings actions are tracked from identification through implementation through verification
- Finance can produce a “verified savings ledger” showing what was committed and what was realized
- Cloud spend decisions are part of product architecture discussions, not just post-deployment reviews
Very few organizations are at the Run stage across all dimensions. The goal is not to claim a maturity level but to identify the specific gaps that are costing the most money and address them in order of financial impact.
The practical maturity assessment
Rather than placing yourself in a stage, answer these five questions honestly:
1. What percentage of your cloud spend can you attribute to a specific team or product? Below 70% is Crawl. 70–90% is Walk. Above 90% with accurate shared cost allocation is Run.
2. How quickly do you detect and route cost anomalies? Weeks (Crawl) → hours (Walk) → automated with owner routing (Run).
3. What percentage of identified savings opportunities are actually implemented? Below 30% is Crawl. 30–70% is Walk. Above 70% with verification is Run.
4. Can you explain last quarter’s cloud cost variance in five minutes without help from engineering? No (Crawl) → with some engineering input (Walk) → yes, with evidence (Run).
5. Do you know the cost to serve one unit of your core product? No (Crawl) → approximately (Walk) → yes, with a trend (Run).
The honest answers to these questions will locate your actual maturity more accurately than any framework classification.
The highest-value next step by stage
If you’re in Crawl: Don’t buy tools yet. Fix tagging first. Define four mandatory tags, enforce them at the account policy level, and retroactively tag your top 20 cost drivers. The visibility this creates will pay for the effort many times over.
If you’re in Walk: Fix the implementation rate. Build a system to track every savings finding from identification to verification. If you’re identifying $500K in annual savings opportunities but only implementing 20%, the problem is not more findings — it’s accountability for following through.
If you’re transitioning to Run: Build unit economics. Work with engineering to define the cost per unit of your core product, measure it monthly, and set an improvement target. This single metric does more to align engineering and finance than any reporting dashboard.
CostDefender is designed for the Walk-to-Run transition: identified savings that are tracked through implementation and verified against actual spend reduction — the closed loop that turns findings into financial results.