Maturity model
A maturity model is only useful if it changes what you do next. Qbits uses a five-level model across five dimensions, scored against observable evidence rather than opinion, and re-scored at defined intervals to show movement.
Levels
| Level | Name | What it looks like | Failure mode |
|---|---|---|---|
| 1 | Initial | Processes are ad hoc and reactive. Success depends on individual effort rather than standard practice. | Heroics mask systemic fragility; knowledge leaves with people. |
| 2 | Repeatable | Basic processes exist but are not standardised or documented. Teams work in inconsistent, siloed ways. | Each team is a separate island; improvements do not spread. |
| 3 | Defined | Processes are documented and standardised across the organisation. Collaboration is beginning; some silos remain. | Documentation drifts from reality; the standard is aspirational. |
| 4 | Managed | Processes are measured. Cross-team collaboration is effective, and automation improves efficiency. | Metrics exist but are not connected to business outcomes. |
| 5 | Optimised | Processes are automated, integrated and continuously improved. The organisation invests in innovation. | Improvement becomes an end in itself; complexity accumulates. |
Most organisations sit at level 2–3 overall while scoring 4 in one dimension — typically automation in a single team — and 1 in another. The spread is more informative than the average.
Dimensions
Process
Are delivery processes defined, followed and measured?
- 1 — No consistent process; each project invents one.
- 3 — A documented SDLC exists and is generally followed.
- 5 — Process is measured, reviewed, and changed based on evidence.
Tools and technology
Is the toolchain fit for purpose and consistently used?
- 1 — Tooling is ad hoc, overlapping, or unmaintained.
- 3 — Standard toolchain exists and is adopted across teams.
- 5 — Toolchain is integrated end to end and evaluated on a cycle.
Team and collaboration
Can teams deliver without organisational friction?
- 1 — Strict silos; handoffs are the primary delivery mechanism.
- 3 — Cross-functional teams exist in places; collaboration is improving.
- 5 — Teams own services end to end; information flows without escalation.
Automation
How much of the delivery path runs without human intervention?
- 1 — Manual builds, manual deployments, manual infrastructure.
- 3 — CI runs; deployment may still be manual or partially scripted.
- 5 — Commit to production is automated, gated and reversible; infrastructure and policy are code.
Governance and compliance
Are controls embedded in the delivery path rather than bolted on?
- 1 — Controls are manual, retrospective, and often bypassed.
- 3 — Controls are documented and applied at release.
- 5 — Controls are automated as policy, evidenced continuously, and versioned.
Scoring rules
- Score the observed practice, not the documented intent. If the standard exists but is routinely bypassed, the score is what actually happens.
- Score each dimension separately. Do not average until the profile is complete — the gaps are the point.
- Require evidence. Every score is supported by at least one artefact: a pipeline, a runbook, a repository, an interview quote, a dashboard.
- Agree the score with the customer. A disputed score is a finding in itself and usually reveals a visibility gap.
- Re-score on a fixed cadence. Quarterly for active engagements; the direction of travel matters as much as the level.
Scorecard
| Dimension | Level | Evidence | Target | Owner |
|---|---|---|---|---|
| Process | ||||
| Tools and technology | ||||
| Team and collaboration | ||||
| Automation | ||||
| Governance and compliance | ||||
| Overall |
From score to roadmap
The gap between current level and target level, weighted by business importance, determines sequencing. A dimension scoring 1 in something the business depends on outranks a dimension scoring 3 in something peripheral.
A typical progression:
- Level 1 → 2: establish source control, a repeatable build, and documented environments. This is foundational; nothing above it holds without it.
- Level 2 → 3: standardise the pipeline across teams, codify infrastructure, and document operating procedures.
- Level 3 → 4: instrument the delivery path and the workload, then manage against the measures.
- Level 4 → 5: automate the controls themselves and close the feedback loop between production signals and delivery decisions.