A maturity model is a diagnostic, not a destination. The useful move is to find the one bottleneck that is actually hurting delivery — then improve that, rather than chasing a level so the annual report can say you have one.
What maturity models are for
Most PMO maturity models describe a ladder: ad hoc practice, repeatable standards, defined process, managed measurement, then optimisation. The labels vary. The idea does not. You are trying to see whether delivery depends on heroics or on a system that would survive the current head of PMO leaving.
Treat the model as a conversation starter with evidence. 'We are level 2 on risk because registers exist but are not reviewed' is useful. 'We are targeting level 4 by December' is a slogan until you name the behaviour that would have to change.
Do not confuse process density with maturity. An organisation can have a 90-page methodology and still run every project as an exception. Maturity is whether agreed practices are used on the work that matters, and whether the results are inspected. Paper is cheap. Habit is not.
Assess without grading your own homework
Pick a simple scale and a small set of practices: intake, planning, risk, change, reporting, benefits, and close. Score each with evidence — artefacts, interviews, and a sample of live projects — not a workshop where everyone votes themselves a B+.
Sample the messy projects, not the showcase. The transformation with the colourful roadmap will look mature. The nine-month integration that has no owner for the operational handover will tell you the truth. If you only assess the poster children, you will write a flattering report and change nothing.
Separate PMO maturity from organisational maturity. The office can be excellent at standards while sponsors still bypass gates. Write both scores. Improving the PMO while leadership ignores it is how you burn out a competent team.
Improve one bottleneck
After the assessment, pick the constraint that is currently wasting the most money or trust. Inconsistent reporting that blinds the portfolio. No intake, so everything is a priority. Risk registers that are never reviewed. One theme, one quarter, one measure of 'better.'
Design a small intervention: a mandatory one-page intake, a monthly portfolio kill-or-keep, a standard gate pack. Resist the urge to launch a 'maturity programme' with seven workstreams. That is how the PMO becomes a project that needs a PMO.
Re-assess the same slice later. If reporting was the bottleneck, look at whether leadership now makes a decision from the pack. A higher model score with the same surprises in steerco is not improvement. It is stationery.
Summary & next steps
Use a maturity model to diagnose, not to decorate. Assess with evidence from ordinary projects, then improve one bottleneck until a behaviour changes.
Next: PMO reporting — the few metrics leadership will actually act on, and the ones that only make the dashboard look employed.