Outcome
By the end of this lesson, you will be able to identify a Walk-stage organization, recognize the savings-decay pattern that defines it, and prescribe the move to weekly Operate cadence that lifts Walk to Run.
| Prerequisites | M5.1.L1: Crawl stage |
| Time | 9 minutes |
1. Concept
Walk is the middle stage. Crawl was seeing without doing. Walk is doing without keeping it up.
These organisations run optimisation sprints, and the sprints work. Real money comes off the bill. Then, over the next six to twelve months, it comes back.
Nothing they are doing is wrong. They are simply not doing it often enough for the wins to stack.
A Walk org looks productive. Sprints happen. Savings show up. The bill goes down. The trap is that nothing in the operating model prevents the savings from drifting away after the sprint is done.
WALK SIGNATURE──────────────────────────────────────────────────────────────────Optimization sprints occur (1-2 per year) ✓ action happensSome teams have schedules running ✓ infrastructure existsRecommendations partially reviewed ⚠ inconsistent triageTag coverage 90-95% ✓ acceptableBudgets per-team exist but irregularly reviewed ⚠ rhythm absentKPIs measured monthly, not weekly ⚠ slow loopCost in engineering's awareness but not workflow ⚠ partial integrationThe signature: action without rhythm. The components exist; they aren’t tied together by a recurring cadence.
The decay pattern
The defining feature of Walk is savings decay: gains achieved in a sprint silently erode in the months that follow.
SPRINT EFFECT example:
Sprint runs Q1 → saves $40K/month at end of Q1 Q2: $35K of savings remain (-12% drift) Q3: $28K of savings remain (-30% from peak) Q4: $22K of savings remain (-45% from peak) Q5: $15K of savings remain (-62% from peak)
By the end of year 1: only 55% of original sprint savings remain; one quarter later, 38%.
The team's reaction is often "we need another sprint": and the cyclerepeats. Each sprint claws back some savings; each post-sprint periodloses some. Net progress is minimal over multi-year horizons.The decay mechanisms are predictable:
DRIFT SOURCE EXAMPLE──────────────────────────────────────────────────────────────────Engineers disable schedules ad-hoc "Need to test something tonight; will re-enable tomorrow" → never does
Recommendations age unactioned New cluster created; right-sizing rec sits open for months
New resources without controls Team launches new workload without tagging or scheduling
Org reorgs reshuffle ownership Team boundary changes; cost owner unclear for weeks
Schedules paused "Just for this incident" → never re-enabled
Budgets unreviewed No one notices the slow climb past commitmentWhy decay happens
The root cause is structural, not technical:
WALK ORG: optimization is project-led, not operate-led
Sprint = project (start date, end date, deliverables) After the project ends, no one is responsible for sustainment
Without weekly Operate cadence: - No one revisits the savings to see if they hold - No one catches when controls slip - No one drives next-quarter optimization - No one updates documentation for new team members - Knowledge from the sprint walks out the door when people leaveThe fix is structural too: shift from project mode to operate mode.
Moving up to Run
The move from Walk to Run is largely about establishing operating rhythm:
1. ESTABLISH WEEKLY OPERATE CADENCE 30-45 minutes per week Attendees: FinOps lead + team cost owners + finance partner Agenda is consistent and short
2. REVIEW THE 5 OPERATE KPIs EVERY WEEK (Covered in T0.M0.2 L4: Operate KPIs) Tag coverage / Schedule coverage / Open rec count / Budget variance / Anomaly count Trends matter more than absolutes
3. TRIAGE RECOMMENDATIONS EVERY WEEK Aim: keep open-rec backlog under 30 days old Anything older than 30 days: re-evaluate or close Cap the backlog at 50 open recs
4. RE-BASELINE BUDGETS QUARTERLY Don't let budgets drift from reality Adjust based on workload changes, growth, optimization gains Document the why in budget notes
5. ANOMALY RESPONSE HAS A CLEAR OWNER On-call rotation or designated FinOps responder <24-hour response time Postmortem after each anomalyThese five practices form the rhythm. The rhythm is what makes Walk → Run real.
Cadence is the discipline
The single most important distinction between Walk and Run:
WALK ORG: monthly review (or none) Monthly is too slow; issues compound between reviews Sprint-shaped work persists Savings decay between sprints
RUN ORG: weekly review (with monthly + quarterly deeper passes) Weekly catches drift before it compounds Continuous incremental optimization Savings compound rather than decayThe cadence difference is what produces markedly better year-over-year outcomes despite the same underlying tools and the same engineering team.
Why Walk orgs stall
A typical Walk-stage org has been in Walk for 1-2 years. The stalling factors:
- Leadership sees savings; thinks the job is done- FinOps team is small; cannot drive operate cadence alone- Engineering team treats cost as "FinOps's problem" still- No one has named the decay pattern explicitly- Tooling supports projects (sprint mode) but not operationsThe naming is often the unlock. Pointing at the decay pattern explicitly (“we saved $40K in Q1 and lost $25K of it by Q3: that’s the Walk pattern”) creates the urgency to establish cadence.
How to diagnose Walk-stage decay
Read the decay pattern from data, not memory. Keep a simple savings log (one row per action: date, resource, monthly saving, still in place yes/no) and re-check it each quarter. Decay shows up as rows flipping to “no”: resources resized back up, schedules removed, commitments left to expire.
WALK SIGNALS TO CHECK EACH QUARTER: Realized savings in your log declining quarter-over-quarter Schedule coverage trending down Open recommendations piling up faster than they close Nobody opening the cost dashboard between sprints (Walk = stale)In ZopNight, two views help. The savings overlay on the Cost Reports trend chart plots reclaimable savings beside spend: when the open opportunity climbs back toward its pre-sprint level, savings are decaying. The Schedules page shows which resources still have a schedule attached.
How to run the Walk → Run transition
Build one canvas for the weekly Operate meeting that shows the 5 KPIs in the order you review them, and keep the agenda in your team wiki. In ZopNight, start from the FinOps dashboard preset (cost trend, cost reports, budget health, top costly resources, recommendations, schedule activity) and clone it into your own dashboard if you need a different set. Bookmark it; one click during the meeting brings the team to the same view. In a cloud-native setup, a saved Cost Explorer or Cost Management view plus a shared spreadsheet does the same job.
2. Demo
A real Walk → Run transition:
WALK ORG: $40K/mo savings achieved Q1 via optimization sprintTIMELINE:
WEEK 0 (end of Q1): $40K/mo savings realizedWEEK 6 (Q2): $36K/mo (-10% drift)WEEK 12 (mid-Q2): $32K/mo (-20%) FinOps lead notices the slopeWEEK 13: Proposes weekly cadence to leadershipWEEK 14: Leadership approves; team cost owners identified; weekly meeting scheduled
WEEK 14-16: Establish + maintain Re-enable schedules that drifted Triage backlog of 47 recommendations Decay pattern explicitly named to engineering First three weekly meetings: 30-min each
WEEK 17-24: Operate rhythm establishes $32K → $38K (re-captured some decay) Weekly KPI dashboard published Cost in engineering's weekly metrics
WEEK 25-52: Compounding effects $38K → $43K (continuous incremental wins) Anomaly response sharpens; <24h typical Tag coverage stabilizes at 95% Forecast accuracy improves to ±8%
END OF YEAR 1 (relative to original sprint): Without weekly Operate (projected): $22K/mo (-45%) With weekly Operate (actual): $43K/mo (+8% beyond original)
Same team. Different cadence. Nearly double the retained savings at year-end; compounds further over multi-year horizons.The naming of the decay pattern, explicitly showing the slope, was the catalyst.
3. Hands-on (5 min)
Assess your org for Walk-stage signals:
WALK SYMPTOMS (check those that apply): □ We've run 1+ optimization sprints in the past year □ Some teams have schedules; coverage is partial □ Tag coverage is 85-95% □ Budgets exist; reviews are monthly (or less) □ Recommendations get triaged but backlog grows □ Cost is in engineering's awareness but not weekly routine □ Realized savings from sprints have declined over time
WALK DIAGNOSIS: 4+ checked = probably Walk
DECAY MEASUREMENT (if you have data): Sprint savings end of Q1: $__________ /mo Same savings end of Q2: $__________ /mo Decay %: _____ %
If decay > 15% in 6 months: Walk-stage decay is real
MOVE-UP PRIORITY: □ Establish weekly Operate cadence (highest leverage) □ Track 5 Operate KPIs □ Recommendation backlog management □ Quarterly budget re-baselineIf you cannot measure decay directly, the meeting cadence question is the quickest diagnostic: “Do we have a weekly meeting where we review cost?” If no, you’re at Walk at best.
4. Knowledge check
Q1
Walk signature: sprint optimizations followed by:
A. Sustained savings forever
B. Bigger savings the next quarter
C. Random outcomes
D. 6-12 month gradual decay
Show answer
Correct: D. Without operate cadence, savings erode through drift; disabled schedules, aged recommendations, new untagged resources, org reshuffles. The decay is gradual enough that it’s not noticed until year-end review.
Q2
Move Walk → Run via:
A. Bigger and more frequent one-time optimization sprints each quarter
B. Weekly Operate cadence with the 5 KPI review and recommendation triage
C. Hire more people
D. More tooling
Show answer
Correct: B. Disciplined recurring practice, not bigger one-time efforts. The cadence is the lever; the rest follows.
Q3
How long is typical decay before it’s noticed?
A. 6-12 months typical
B. Days
C. Years
D. Immediately, within the same month
Show answer
Correct: A. Subtle and gradual until visible at year-end review. The slow timeline is what makes Walk persistent: each quarter looks “mostly fine” until you compare to the original sprint baseline.
5. Apply
Establish weekly Operate cadence; track the 5 KPIs on one bookmarked view (in ZopNight, a dashboard built from the FinOps preset). Keep the agenda and the savings log in your team wiki.
If you’re already running optimization sprints, the win is preserving what you’ve built: the Operate cadence is much cheaper than another sprint and produces compounding rather than one-time gains.
Related lessons
- L1: Crawl: visibility-only orgs
- L3: Run: operate cadence compounds (next)
- L4: Maturity anti-patterns
- L5: 90-day move-up plan
- T0.M0.2.L4: The 5 Operate KPIs
Glossary terms touched
Crawl-Walk-Run · Savings decay · Operate cadence