Outcome
By the end of this lesson, you will be able to self-assess an organization against the three FinOps maturity stages and draft a 90-day move-up plan.
| Tier | Operator |
| JTBD | ”Tell me where my org actually sits on the maturity curve, no flattery.” |
| Personas | FinOps Analyst · Engineering Leader · Finance Partner |
| Prerequisites | L1-L4 |
| Time | 10 minutes |
| Bloom verb | Self-assess (Evaluate) and Draft (Create) |
1. Concept
Practices mature in three recognisable stages, named Crawl, Walk and Run.
They are not levels you unlock in order. They are names for where you actually are, which is useful mainly because it makes the next conversation honest.
STAGE DEFINING TRAIT TYPICAL TIME IN STAGE─────────────────────────────────────────────────────────────────────CRAWL Spend visibility exists, action does not 6-18 monthsWALK Optimization happens, decay still wins 12-24 monthsRUN Operate cadence compounds savings ongoingAlmost everyone stays in Crawl longer than they believe, and reaches Run later than they planned.
Crawl
What it looks like: the company can produce a bill but cannot reliably act on it.
The data exists, in spreadsheets or a dashboard or a tool. The bill gets read and reports go out. What is missing is the ability to change anything. Nothing is scheduled off. Recommendations pile up unreviewed. Fewer than 70% of resources carry tags, and no team has a budget anyone enforces.
The usual mistakes here:
- A cost team owns the bill but cannot get any team to turn something off
- A monthly report exists and is understood to be “for finance”
- Engineers cannot say what their own team spends
- Commitments get bought one at a time, without anyone modelling the floor first
The giveaway: dashboards are being opened often, and almost no decisions are being made each week.
To move up: give each team ownership of its own cost. Get tags onto more than 90% of resources. Put at least one non-production environment on a schedule. Publish one weekly set of numbers.
Walk
What it looks like: savings happen, then quietly unwind unless somebody keeps pushing.
There is a rhythm of checks, but it runs monthly rather than weekly. Schedules exist and drift. Recommendations get read but not consistently acted on. Some teams have budgets and some do not. Tags cover 90 to 95% of resources.
Anti-patterns at this stage:
- Sprint-driven optimizations followed by quiet six-month decay
- Reservations bought before steady-state floor is known
- Some teams excel, others ignore: no consistency
- KPIs measured monthly, not weekly
Signature symptom: the same recommendation appears, gets dismissed, appears again two months later.
To move up: weekly Operate cadence. Five KPIs reviewed every week. Every team has a budget and reads it. Anomalies have an owner. Commitments calibrated to post-schedule floor.
Run
Defining trait: The Operate cadence compounds. New recommendations are reviewed weekly and triaged within days. Savings persist because the cadence persists. Unit economics is a primary KPI alongside total spend. Tag coverage is above 98%. Every team has a budget. Anomalies are root-caused inside 24 hours.
Anti-patterns at this stage:
- Treating Run as a finish line (“we made it, we can relax”)
- Letting the Operate cadence skip during major events without scheduling a make-up
- Failing to refresh the framework as new clouds or services are adopted
Signature symptom: the avoidable-spend KPI is consistently below 5% and unit economics tracks business growth.
To stay at Run: never drop the weekly cadence. Re-baseline quarterly. Bring in new clouds / new tooling under the same framework.
Stage is per-capability, not per-organization
The FinOps Foundation’s full framework rates maturity per capability (there are 22 of them). An organization can be at Run on Reporting & Analytics, Walk on Anomaly Management, and Crawl on Forecasting all at once. The averaged stage is less useful than the per-capability heat map.
For a self-assessment, the practical model is: pick the four to six capabilities most relevant to the next quarter’s goals. Score each. Pick the worst two. The 90-day plan focuses on moving those two from Crawl to Walk, or Walk to Run.
The 90-day move-up plan
WEEK ACTION─────────────────────────────────────────────────────────────1-2 Self-assess (the table above). Pick two worst capabilities.3-4 Publish the 90-day plan. Establish the weekly cadence.5-8 Execute on the two worst capabilities (concrete actions).9-10 Re-baseline. Are the two worst now Walk instead of Crawl?11-12 Identify the next two worst. Repeat.─────────────────────────────────────────────────────────────The pattern is narrow focus + recurring cadence. Not broad sweep + one-time push.
2. Demo
A real self-assessment from an Engineering-Leader-led FinOps practice:
CAPABILITY STAGE EVIDENCE─────────────────────────────────────────────────────────────────Data Ingestion Walk Billing data daily, cost source labelledAllocation Walk Team tagging 91% coverageReporting & Analytics Walk Weekly KPI dashboard existsAnomaly Management Crawl Anomalies detected but no ownerForecasting Crawl Quarterly forecast, no accuracy trackingBudgeting Walk Per-team budgets, alerts wiredUnit Economics Crawl Defined but not trackedArchitecting / Workload Place. WalkUsage Optimization Walk Schedules + recs, lever 1+2 activeRate Optimization Crawl Commitments not modelled─────────────────────────────────────────────────────────────────WEAKEST TWO: Anomaly Management, Forecasting90-DAY PLAN: assign anomaly owner per team; add forecast accuracy trackingTwo clear targets. Twelve weeks to move them from Crawl to Walk.
3. Hands-on (7 min)
Pick six capabilities you can confidently score:
1. Data Ingestion Crawl / Walk / Run2. Allocation Crawl / Walk / Run3. Reporting & Analytics Crawl / Walk / Run4. Anomaly Management Crawl / Walk / Run5. Budgeting Crawl / Walk / Run6. Usage Optimization Crawl / Walk / RunFor each Crawl: what is the single highest-impact thing to do in the next 30 days that would move it to Walk? Write it. The 90-day plan is the consolidation of the answers.
4. Knowledge check
Q1
A team can pull the bill, has good dashboards, but cannot get any team to act on the data. The team is most likely at:
A. Crawl stage on every capability
B. Walk
C. Walk on Inform, Crawl on Optimize
D. Run
Show answer
Correct: C. Per-capability assessment is the right model. Strong Inform with weak Optimize is the textbook profile of an organization stuck before Walk.
Q2
A team reports: “We did an optimization sprint, saved $40K monthly, but six months later only $22K of those savings remain.” Most likely stage:
A. Crawl
B. Walk
C. Run
D. Beyond Run
Show answer
Correct: B. This is the classic Walk symptom: optimization happens, savings decay. Run organizations do not see this pattern because the Operate cadence holds the line.
Q3
A 90-day move-up plan should:
A. Address all 22 capabilities at once
B. Hire a consultant first
C. Wait until the next quarter before beginning on the work
D. Focus on the two worst capabilities and move them up one stage
Show answer
Correct: D. Narrow focus, recurring cadence. Trying to fix all 22 capabilities at once is the failure mode that produces nothing moving.
5. Apply
ZopNight does not (yet) ship a self-assessment widget. The capability scoring is a manual exercise done annually or quarterly.
What ZopNight does support directly:
- Crawl → Walk moves on Inform (Reports surface) and Allocation (Tag attribution)
- Walk → Run moves on Anomaly Management (15-min detection) and Budgeting (per-team budgets with thresholds)
- Walk → Run moves on Usage Optimization (recommendations + auto-remediation + scheduling)
The T4: FinOps Mastery track expands on the maturity ladder with per-capability deep dives.
Related lessons
Glossary terms touched
Crawl-Walk-Run · Maturity model · Per-capability assessment