The Data Center Maturity Model (DCMM) is the model published in 2011 by The Green Grid, the international consortium for the efficiency of IT infrastructures, to measure a data centre's efficiency and sustainability. It assesses the infrastructure on a scale of six levels (0–5) and across two broad dimensions — the facility (power, cooling, building, management) and IT (compute, storage, network) — with a simple principle: each additional step in maturity brings more efficiency and sustainability, in return for an investment of money, time and resources. The outcome is not a score but a roadmap: knowing where you are today, to decide where to go tomorrow.
Six levels, one direction
0 · Starting point
No measurements and no plan: nobody knows how much is consumed or how much is used, and action is taken only when something stops.
1 · Partial best practices
First measurements and first interventions, but patchy: some things are optimised, the rest coasts along.
2 · Best practices
Established practices applied consistently: you measure, you manage, you document. This is the threshold of good management.
3–4 · Beyond current practice
Advanced optimisation and automation: decisions follow defined metrics and policies, not habits.
5 · Visionary
The model's horizon: infrastructures that adapt to the load by themselves, low-emission energy, consumption proportional to the work done.
What the model looks at
Facility
- Power: efficiency of the electrical chain, UPS, low-emission generation
- Cooling: free cooling and economisation, set-points, airflow management
- Resilience and building: mapping resilience to actual need, lighting, building envelope
- Management metrics: PUE, heat reuse (ERF), carbon footprint (CUE) and water footprint (WUE)
IT
- Compute: actual server utilisation, virtualisation, power management, TCO-driven refresh
- Storage: deduplication, data tiering, capacity managed on actual need
- Network: bandwidth utilisation, consolidation, components switched off when not needed
- The IT around it: sizing, efficient power supplies, service catalogue, incentives for efficient behaviour, e-waste and procurement
A few thresholds, to be concrete
The model is not made of adjectives but of measurable criteria. A few examples, from level 1 to level 5:
90% → 96%
efficiency of the critical electrical chain
0.5 → 0.05
annual contribution of cooling to PUE
>20% → >60%
average monthly server utilisation (from level 2)
plan → 0.5
waste heat reuse (ERF factor)
You don't need to be a hyperscaler
The DCMM was created for large data centres, but its questions apply just the same to a server room with three racks: how much does cooling consume compared with the IT? How hard do the servers actually work? Is cold data sitting on expensive storage? Is there a plan, or do you act on failure? That is why we use it as a reading grid in smaller environments too — indeed, especially there:
- Measure before investing. The first step often costs next to nothing: a power meter, utilisation monitoring, even an approximate PUE. Without numbers, every investment is a gamble.
- The most profitable jumps are the first ones. Moving from levels 0–1 to level 2 is largely a matter of low-cost interventions: airflow management (hot/cold aisles, blanking panels), correct set-points, virtualisation and consolidation, switching off what is not needed, deduplicating data.
- The right level is not 5. It is the one consistent with your business: the roadmap is there to choose where climbing is worth it — and where it is not.
- Scale doesn't matter; method does. We have tested it on ourselves: our Green Data Centre — free cooling, operation at ~28 °C, 100% renewable energy — applies the principles of the model's upper levels in a facility that is anything but a hyperscaler.
How it works
- Survey and assessment of the current state across the model's facility and IT dimensions
- Positioning on the maturity levels, area by area, and identification of the gaps
- An evolution roadmap with priorities, effort and expected benefits — sized to your scale
- The metrics (PUE, utilisation, capacity) feed into continuous monitoring: maturity is maintained, not certified once and for all
- Integration with methodologies and standards: ITIL, service-oriented architectures
Benefits
- Investment directed where it produces the most value
- Planned evolution instead of chasing emergencies
- A shared number to discuss with management and ownership: the starting level and the one to reach
- The ideal starting point for infrastructure renewal or continuity projects
The reference model is the ‘Data Center Maturity Model’ v1 (2011) by The Green Grid; the thresholds quoted are the model's own, while the reading for smaller environments comes from our experience in the field.
