Refine & scale
Maturity model
A scale from "ad hoc" to "optimized" — applied to your own domains. Measure periodically, grow visibly, invest where the pain is.
Updated May 18, 2026
Refine · 6.4
Why this matters now
Without a model, teams talk in general terms (“it works”, “it could be better”). With a model, they talk in levels: “ticketing sits at 3, we aim for 4 within 6 months, and that requires X and Y.” That turns investment conversations from “trust us” to “look at the score”.
The five levels
1 — Ad hoc
No fixed process. Different every time. Outcome depends on who's doing it.
2 — Managed
Process exists, is followed, but not yet consistent and without measurable KPIs.
3 — Standardized
Process is documented, KPIs measured, training in place. Predictable for 80% of cases.
4 — Measured & steered
KPIs actively steer adjustments. Decisions based on data. Continuous-improvement log active.
5 — Optimized
AI & automation support, self-correcting, benchmark level. Truly differentiating.
Domains to score
Incident & service management
SLA compliance, routing, knowledge-base usage, escalation discipline.
Workplace & reservations
Occupancy measurement, no-show management, space optimization.
Visitor management
Pre-registration, hospitality, compliance, evacuation readiness.
Products & chargeback
Up-to-date catalogue, ERP integration, cost-center mapping, audit conformity.
Data quality & classification
Completeness, consistency, free-text share.
AI maturity
Number of use cases, average level, trust score.
Adoption & satisfaction
Active users, CSAT, NPS, opt-out reasons.
Governance & ownership
Roles filled, RFC flow used, audits documented.
Key questions
- 1Which domains do you actively measure? Start with 5-8, score each quarter with both the service owner and the sponsor.
- 2Who scores — self-evaluation, peer review, external? A combination gives better accuracy.
- 3Evidence per level — which artefacts (documents, dashboards, logs) prove you're really at that level?
- 4Target level per domain — not everything needs to go to 5. Managed (2) or Standardized (3) is enough for some domains.
- 5Next step — if you're at 2 and aim for 3: what concretely has to happen? Documentation? Training? KPIs? Tooling?
- 6Frequency — quarterly seems ideal. Monthly = noise, yearly = too slow to steer.
- 7Benchmark — do you compare against external data (similar organizations, sector averages) or only yourself over time?
- 8Communication — how do you present scores to the steering committee and broader audience? A spider chart works visually.
Template — Score sheet (quarterly)
| Domain | Score now | Score 6m ago | Target 12m | Evidence | Next step |
|---|---|---|---|---|---|
| Incident management | 3 — Standardized | 2 | 4 | SLA report, workflows, KB | Activate continuous-improvement log |
| Workplace & reservations | 3 | 2 | 4 | Occupancy dashboard, no-show policy | Connect sensor data |
| Visitor management | 2 — Managed | 2 | 3 | Pre-registration 60% | Make mandatory, fine-tune kiosk flow |
| Products & chargeback | 3 | 2 | 3 | ERP integration, monthly export | Target reached — maintain |
| Data quality | 2 | 1 | 3 | 75% classification filled | Mandatory fields, AI suggestions |
| AI maturity | 2 | 1 | 4 | 3 use cases live, 1 at level 2 | Promote no-show detection to 3 |
| Adoption & satisfaction | 4 — Measured & steered | 3 | 4 | CSAT > 4.2 / adoption 91% | Maintain, don't slacken |
| Governance & ownership | 3 | 2 | 3 | RFC flow active | Expand audit trail |