Outcome
By the end of this lesson, you will be able to identify a Crawl-stage organization from its symptoms, diagnose the specific gaps holding it at Crawl, and prescribe the four move-up actions that lift an org from Crawl to Walk.
| Prerequisites | T0: Foundations |
| Time | 9 minutes |
1. Concept
The FinOps Foundation names three stages: Crawl, Walk and Run. Every organisation starts at the first. Strictly, the Framework scores maturity per capability, so an org can be Run on scheduling and Crawl on unit economics; this module uses the stage names as shorthand for an org’s dominant pattern.
Crawl is seeing without doing. The dashboards exist. The reports go out. Everybody knows the number. And nothing happens as a result: recommendations sit unread, tickets pile up, and the FinOps team does all the work itself because nobody else will.
A Crawl org is not a failing org. It is a typical org early in its cost-discipline journey. The challenge is moving past it deliberately, not pretending it doesn’t exist.
Crawl symptoms
SIGNAL WHAT IT INDICATES──────────────────────────────────────────────────────────────────FinOps team owns the bill No team-level ownership; cost is "finance's problem"
Cost report is a monthly PDF Engineering doesn't read; decisions don't happen in the PDF format
Tag coverage <70% Most spend Unattributed; attribution arguments dominate cost conversations
Recommendations sit unreviewed Backlog grows; the highest- value recs are old and stale
Reservations bought ad-hoc No commitment modeling; often over-committed or under-committed
Engineers don't know team cost Cost not part of the engineering team's daily decisions
Weekly meetings have cost as agenda But action items rarelyitem but never as action item assigned or completed
Pre-merge cost estimation absent Engineers don't see cost impact of their code changesThe cluster is recognizable: information exists in places that don’t change behavior. The org has the tools; it lacks the operating model.
Why orgs stay at Crawl
Crawl is not a stop the org consciously chooses. It is where many orgs land by default because the conditions to advance are absent:
CONDITION ABSENT WHY IT MATTERS──────────────────────────────────────────────────────────────────Team-level cost accountability Engineering teams won't act on cost unless they own it
Operating cadence (weekly review) Without recurring rhythm, cost conversations are ad-hoc and forgotten
Tagging discipline Attribution fights consume energy that should go to optimization
FinOps + engineering collaboration FinOps in isolation recommends; engineering in isolation ignores
Cost-aware tooling in engineering If cost isn't in the IDE,workflows PR, dashboard engineers use, cost is out-of-bandA Crawl org can stay at Crawl for years if no one names the gap and drives the move-up.
The four move-up actions (Crawl → Walk)
In priority order, the actions that consistently move orgs to Walk:
1. ESTABLISH TEAM-LEVEL COST OWNERSHIP - Each team has a dashboard showing their resources and cost - Each team has a designated cost owner (team lead or assigned practitioner) - Cost appears in team's regular metrics - First step because everything else depends on this
2. GET TAG COVERAGE ABOVE 90% - Roll out Smart Tags; accept the derived values that identify an owner - Manually clean up high-spend Unattributed resources - Enforce tag policy in IaC (Terraform pre-merge checks) - Closes the attribution-fight loop
3. SCHEDULE AT LEAST ONE NON-PROD ENVIRONMENT - Pick a dev/staging environment; schedule it on/off - Demonstrate that operational discipline produces real savings - Builds the muscle for broader scheduling - Often saves 30-50% of non-prod cost
4. PUBLISH A WEEKLY KPI DASHBOARD - Daily or weekly cadence beats monthly - Key signals: total spend, team variance, tag coverage, recommendation backlog size - Distributed to engineering leadership + FinOps + finance - Establishes the rhythmThese four actions, executed deliberately, move an org from Crawl to Walk in 60-90 days. Trying to do all four simultaneously usually means none of them lands; sequencing matters.
What NOT to do at Crawl
The temptation at Crawl is to skip ahead: buy commitments, build a Sankey, hire a FinOps team. Each can wait until the foundation is in place:
TEMPTATION WHY NOT YET──────────────────────────────────────────────────────────────────Buy reservations / Savings Plans Without commitment modeling you'll under- or over-commit
Build elaborate dashboards No one will look until ownership exists
Hire a FinOps consultant team Consultants can advise but cannot replace the operating model your org needs to build
Roll out auto-remediation Risk of breaking things without the operating cadence to monitor results
Quarterly cost reviews Without weekly rhythm, quarterly conversations lack continuityThe right approach: focus on the four move-up actions first. Defer everything else.
How to self-assess Crawl
Score yourself against the FinOps Foundation maturity model: rate each capability Crawl, Walk or Run, or 0-5 if you want finer grain. Do it as a 45-minute working session with FinOps, one engineering lead and one finance partner, and write down the evidence behind each score. A score with no evidence is a guess. For Crawl-stage orgs, the gaps almost always include team ownership, tagging, and operating cadence.
SELF-ASSESSMENT EXAMPLE (Crawl): Visibility: 4/5 (dashboards exist; data is there) Ownership: 1/5 (FinOps alone; no team owners) Operating model: 1/5 (no weekly cadence; monthly PDFs) Tagging: 2/5 (60% coverage; no policy) Commitments: 0/5 (no model; ad-hoc buys) Optimization: 2/5 (recs exist; not actioned)
OVERALL: CrawlTOP 3 GAPS: Ownership, Operating model, TaggingFIRST ACTION: Team ownershipPull the evidence from the cost data you already have. Your provider’s billing console (AWS Cost Explorer, Azure Cost Management, GCP Billing reports) shows how much spend carries an owner or cost-center tag. In ZopNight, the Tag Coverage card on Cost Reports, Tags tab, gives the tagging number, Cost Reports, Teams tab, shows how much spend has a team owner, and the commitment coverage card on Cost Reports shows how much of the bill runs on commitments. ZopNight does not compute a maturity score; the scoring is yours.
For mature orgs that want to identify outlier teams, run the same assessment per team: a Run-stage org may have one Crawl-stage team that needs targeted support.
The Crawl mindset shift
The hardest part of Crawl → Walk is not technical; it is mindset. The org has to stop treating cost as “finance’s problem” or “FinOps’s problem” and start treating it as “everyone’s problem, with team-level ownership.” This shift is more cultural than technical.
CRAWL MINDSET WALK MINDSET──────────────────────────────────────────────────────────────────"FinOps will tell us if cost is bad" "We track our cost weekly""Tag drift isn't important" "Tagging is part of our PR""Reports come from finance" "We read our team's report""Recommendations are FinOps's job" "We action our own recs""We can wait for end-of-month" "We catch issues this week"Move-up actions land best when the mindset is shifting in parallel; the team is becoming the owner, and the tools support the ownership.
2. Demo
A team’s actual Crawl → Walk transition over 12 weeks:
WEEK 0: BASELINE Tag coverage: 62% Monthly cost: $180K; report nobody reads Schedules: 0 (everything always-on) FinOps team owns the bill alone Weekly cadence: none Recommendations open: 47; oldest 4 months old Engineering knowledge of own team cost: low
WEEK 1-4: TEAM OWNERSHIP + TAGGING Action 1: Designated team cost owners (one for each of the 5 teams); per-team dashboards published; teams trained Action 2: Smart Tags run; accept the derived values
Tag coverage: 62% → 84% Team owners identified: 5/5 Per-team dashboards: 5 published Awareness: teams now see their cost weekly
WEEK 4-8: SCHEDULING + RHYTHM Action 3: Schedule on dev-platform group (47 resources) Action 4: Weekly KPI dashboard reviewed in a new weekly meeting (FinOps + team leads)
Tag coverage: 84% → 91% Schedules deployed: 1 environment Schedule savings: $4K/month Weekly meeting: established; 30 min recurring Recommendations actioned: 12 closed in 4 weeks
WEEK 8-12: EXPANSION + STABILIZATION Action 3 (extended): Schedules on 2 more environments (staging, sandbox) Action 4 (extended): Weekly KPI email to engineering leadership
Tag coverage: 91% → 94% Schedules deployed: 3 environments Total schedule savings: $12K/month Weekly meeting: 30 min; team leads attend Recommendations actioned: 26 closed; backlog at 21 Engineer awareness of cost: high
WEEK 12: WALK STAGE ACHIEVED Tag coverage: 94% (above the 90% threshold) Each team has their cost numbers Weekly cadence established and sticky Operating discipline demonstrated Next stage: M5.1.L2 (Walk)8-12 weeks for a deliberate transition. The four actions, sequenced, deliver the maturity stage shift.
3. Hands-on (5 min)
Self-assess your org on the Crawl symptoms:
SYMPTOMS PRESENT (check those that apply): □ FinOps team owns the bill alone □ Monthly cost report (PDF or similar) □ Tag coverage <70% □ Recommendations sit unreviewed for weeks □ Reservations bought ad-hoc (no model) □ Engineers don't know their team's cost □ Weekly meetings don't include cost □ Pre-merge cost estimation not in place
CRAWL-STAGE DIAGNOSIS: How many checked? _____ of 8 4+ checked: probably Crawl
TOP 3 GAPS to close (from your checks): 1. __________ 2. __________ 3. __________
FIRST MOVE-UP ACTION: (Pick from the four; usually team ownership goes first) Action: __________ Owner: __________ Timeline: __________If you checked fewer than 2 symptoms, you’re past Crawl. Move to L2 (Walk).
4. Knowledge check
Q1
Crawl is defined by:
A. No data at all
B. Mature operations throughout
C. Just commitments
D. Visibility without action
Show answer
Correct: D. The org has dashboards and reports, but consistently cannot get teams to act on the findings. The information exists in places that don’t change behavior. Visibility vs action is the Crawl signature. The fix is operating model, not more data.
Q2
The first move-up action from Crawl is typically:
A. Buy reservations
B. Establish team-level cost ownership
C. Hiring an external FinOps consultant
D. Quarterly cost reviews
Show answer
Correct: B. Each team gets a dashboard, a designated cost owner, and cost in their regular metrics. Everything else depends on this; without ownership, dashboards go unread, recs unactioned, schedules un-built.
Q3
Typical Crawl → Walk timeline:
A. 1 day
B. 1 year
C. 2 weeks of focused work
D. 60-90 days of deliberate work across the four move-up actions
Show answer
Correct: D. Faster than that is usually superficial; slower than that means the work isn’t focused.
5. Apply
Run the self-assessment above with FinOps, one engineering lead and one finance partner. Score each capability, write the evidence next to each score, and pick the top 3 gaps. Use your billing console or ZopNight’s Cost Reports (Teams and Tags tabs) as the evidence for the ownership and tagging scores.
For new orgs starting at Crawl, schedule a kick-off meeting where the four move-up actions are owned and scoped. 60-90 days from kick-off to Walk is achievable with consistent focus.
Related lessons
- L2: Walk: weekly Operate cadence (next)
- L3: Run: forecasted and automated
- L4: Maturity anti-patterns
- L5: 90-day move-up plan
- T4.M4.5.L4: Tag coverage
Glossary terms touched
Crawl-Walk-Run · FinOps Foundation · Cost ownership · Operate cadence