Why Maturity Models Matter
FinOps maturity models exist because "doing FinOps" means radically different things at different organizations. A startup with $50k/month in cloud spend and a single engineer tracking costs in a spreadsheet is doing FinOps. So is a Fortune 500 company with a dedicated FinOps platform, real-time anomaly detection, and automated rightsizing across 50,000 instances. The practices that are right for one are wildly inappropriate for the other.
Maturity models provide a common vocabulary for describing where you are and a roadmap for where to go next. They prevent organizations from either over-investing in capabilities they're not ready to leverage or under-investing in foundations they desperately need.
The IFO4 FinOps Maturity Framework uses four stages: Crawl, Walk, Run, and Optimize. Each stage is defined not by tools or headcount, but by organizational capabilities - what the organization can consistently do, not just what it aspires to do.
Stage 1: Crawl - Visibility
The defining question: Can you answer "how much are we spending on cloud, and who is spending it?"
Most organizations enter FinOps in the Crawl stage, often triggered by a budget overrun or a sudden executive interest in cloud costs. The primary challenge at this stage is visibility: getting cost data into a usable form and mapping it to the people and teams responsible for the spend.
Key capabilities at Crawl:
- Cloud billing data is accessible and centralized (Cost Explorer, GCP Billing, Azure Cost Management) - Resources are tagged with at least team and environment - Cost anomalies are detectable within 72 hours (not necessarily automated) - There is at least one person with "FinOps" as a defined responsibility - Engineering teams are aware that their cloud usage has a cost
What Crawl organizations typically struggle with:
Tag coverage below 80%, cost data that lags 24-48 hours, no chargebacks (all costs absorbed by a central IT budget), and limited awareness among developers of the cost implications of architectural choices.
The Crawl exit criteria: You can consistently answer "which team spent what, in which service, this month" within 48 hours, with 90%+ cost coverage.
Stage 2: Walk - Control
The defining question: Can you act on what you see - rightsizing, reservation coverage, and waste elimination?
Walk-stage organizations have achieved basic visibility and are now building the processes to act on it. This is where most FinOps optimization work happens: identifying and eliminating obvious waste, building reservation portfolios, and establishing the workflows that connect cost insights to engineering action.
Key capabilities at Walk:
- Rightsizing recommendations are reviewed and acted on within a defined SLA (typically monthly) - Reserved Instance or Savings Plan coverage exceeds 60% of eligible spend - Chargeback or showback reports are distributed to cost center owners monthly - FinOps reviews are included in sprint planning or architecture review processes - Unit cost metrics (cost per request, cost per customer) are defined even if not yet widely tracked
What Walk organizations typically struggle with:
Reservation management is manual and reactive (purchasing RIs based on current usage rather than forecast), chargeback is showback (teams see the data but feel no financial accountability), and optimization is driven by FinOps team recommendations rather than embedded in engineering workflows.
The Walk exit criteria: >70% reservation coverage, <15% identified waste, and at least three engineering teams have cloud cost as a factor in their regular sprint review.
Stage 3: Run - Governance
The defining question: Is cost efficiency embedded in how engineering makes decisions, not just how FinOps reports on them?
Run-stage organizations have moved FinOps from a reporting function to a governance function. Cost considerations are embedded in architecture review boards, deployment pipelines, and capacity planning processes. Engineers consider cost as a first-class constraint alongside performance and reliability.
Key capabilities at Run:
- Cost-per-unit metrics are tracked in production dashboards alongside latency and error rate - Architecture decisions include a FinOps review for significant new services or migrations - Anomaly detection is automated with P1/P2/P3 severity classification and owner notification - Reserved capacity is managed through a forecasting model, not just current utilization - Chargebacks create real financial accountability (costs appear in team budgets or OKRs) - FinOps has executive sponsorship at the VP or C-suite level
What Run organizations typically struggle with:
Multi-cloud cost management (each provider has different tools and data models), SaaS spend governance (often outside the FinOps team's purview), and connecting cloud optimization ROI to business outcomes in terms that resonate with the board.
The Run exit criteria: >85% reservation coverage, <8% waste, anomaly MTTR under 4 hours, and cost-per-unit metrics available for every tier-1 service.
Stage 4: Optimize - Survivability
The defining question: Are you continuously improving efficiency at scale, across all technology spend categories?
Optimize-stage organizations represent the top tier of FinOps maturity. They have moved beyond cloud-native cost management to manage all technology spend with financial discipline - including SaaS, on-premises, data center, and AI/ML infrastructure. Optimization is largely automated, with human oversight focused on strategic decisions rather than tactical waste elimination.
Key capabilities at Optimize:
- FinOps scope extends beyond cloud to SaaS management, licensing optimization, and data center rationalization - Optimization is largely automated: rightsizing happens continuously without manual intervention - AI/ML spend has dedicated governance including per-experiment cost tracking and model efficiency metrics - Technology total cost of ownership (TCO) analysis informs build-vs-buy and cloud-vs-on-prem decisions - FinOps is a competitive advantage, not just a cost control mechanism - efficiency enables faster product investment
What distinguishes Optimize from Run:
The shift from reactive to predictive. Optimize-stage organizations don't just respond to cost anomalies; they forecast them. They don't just optimize existing infrastructure; they design new systems with efficiency as a primary constraint from day one.
Advancing Your Maturity: Practical Next Steps
From Crawl to Walk:
1. Achieve >90% tag coverage - without this, Walk capabilities are built on sand 2. Run your first Reserved Instance purchase, even a small one, to build internal expertise 3. Establish monthly cost reviews with at least two engineering teams 4. Define your first unit cost metric and start tracking it
From Walk to Run:
1. Automate rightsizing recommendations and establish a monthly action SLA with engineering 2. Build a forecasting model for RI/SP purchases 3. Implement automated anomaly detection with owner routing 4. Achieve real chargeback (not just showback) for at least one business unit
From Run to Optimize:
1. Expand scope to SaaS management 2. Implement AI/ML cost governance 3. Integrate FinOps metrics into executive business reviews alongside revenue and margin 4. Build an internal FinOps center of excellence to scale practices across the organization
Maturity is not about perfection at any single stage before moving to the next. Many organizations operate at Walk for most capabilities while exploring Run-level practices in specific areas. The model is a guide, not a checklist. Use it to identify your highest-impact next steps, not to grade yourself against an abstract ideal.