What is PMO maturity?

The gap between levels is not even

Moving from level one to level two is largely a matter of writing things down. A template library, a reporting cadence, an agreed escalation path. It takes effort but it takes no courage, and most organisations that decide to do it succeed.

Moving from level two to level three is the hard one. It is where most PMOs stall, and it is the transition that separates a programme office that helps from one that holds.

The reason is that level three requires the practice to apply when applying it is inconvenient. A stage gate that can be waived under schedule pressure is not a gate. A risk threshold that is quietly raised when the number goes past it is not a threshold. At level two, the control exists. At level three, the control binds. Nothing about the documentation changes between the two. What changes is whether the organisation is willing to be told something it does not want to hear.

That is why maturity work so often fails when it is run as a documentation exercise. The artefacts were never the constraint.

The test that separates level two from level three

Ask what happens to the process when a programme goes red.

At level two the templates are abandoned the moment delivery is under threat, because they were a reporting convenience rather than a control. The status pack gets thinner. The risk log stops being updated. The governance forum turns into a recovery meeting and never turns back. At level three the practice holds through the red period, and the reporting gets more honest rather than less.

There is a simpler version of the same test. Take the person who built the PMO out of the room for a month. If the cadence, the escalation path and the decision record survive without them, the practice is defined. If any of the three degrades, the practice was personal.

Both tests measure the same thing from different angles. Whether the discipline is owned by the organisation or by an individual.

Why most PMO maturity assessments overstate the result

Three failure modes recur, and together they explain why so many organisations hold a flattering assessment and a struggling portfolio at the same time.

The assessment scores existence rather than use. A question that asks whether a benefits register is maintained will be answered yes if the file exists. The useful question is when the register was last challenged, and by whom, and what changed as a result. Existence is cheap. Use is not.

The assessment is completed by the party being assessed. A PMO scoring its own maturity is being asked to document its own shortfall, usually in a period when its funding is under review. The bias is structural rather than dishonest, and it is not removed by asking people to be candid. It is removed by evidence.

The assessment is run once, while the portfolio is calm. Maturity is a claim about behaviour under load. Assessing it during a quiet quarter measures the intention rather than the practice.

An assessment worth acting on tests the same claims against artefacts that already exist. Decision records, estimate history, the last three escalations. Those cannot be improved on the day of the assessment, which is precisely what makes them useful.

Maturity is not the objective. Fit is.

Level five is not the target for every organisation, and pursuing it is often waste.

A PMO carrying three modest projects a year does not need quantitative portfolio management. Building it will consume more than it returns, and the overhead will be visible to everyone except the people who built it. The right question is not how mature the PMO is. It is whether the PMO is mature enough for the portfolio it carries.

A level two PMO against a forty million dollar ERP replacement is under-built, and the shortfall will surface at cutover when it is expensive to fix. A level four PMO against a maintenance portfolio is overhead dressed as rigour.

Maturity is a means. The portfolio sets the requirement.

How to assess PMO maturity honestly

Assess against evidence rather than assertion. Four sources carry most of the signal, and all four already exist in any organisation running programmes.

The decision log. Read the last twenty decisions. Each should carry the options considered, the rationale, an owner and a date. If the log records outcomes without rationale, decisions are being made somewhere other than the forum that claims to make them.

Estimate variance. Compare the last five programmes against their approved baselines. A PMO that cannot produce the comparison is below level four by definition. A PMO that can produce it and has never used it to change an estimating practice is at level four and not moving.

The escalation trail. Take the last three escalations and follow them end to end. How long between the issue being known at working level and being visible at governance level? A long gap is not a communication problem. It is a signal that the escalation path costs the person using it something.

The last red programme. Read what the reporting said in the two months before it turned red. If the pack said amber throughout and then went red without an intervening event, the reporting was managing perception rather than describing reality.

Four sources, all retrospective, none of which can be prepared for. That is the point.

What actually moves a PMO up a level

Three things move a PMO, and none of them is a framework.

The first is a mandate that survives contact with a senior stakeholder. A PMO that can be overruled by whoever is most senior in the room on the day cannot hold a control, and no amount of process design will change that. The mandate is granted, not earned, and it is granted by the executive who is willing to be told their programme is in trouble.

The second is a shorter distance between the working level and the governance forum. Most reporting failures are not dishonesty. They are the accumulated effect of four layers of summarising, each of which rounds slightly towards comfort. Removing a layer does more for reporting quality than rewriting the template.

The third is a reason to keep the practice when it is inconvenient. That reason is almost always a recent, visible instance of the practice being right. A stage gate that stopped a bad go-live is worth more to the next twelve months of governance than any maturity roadmap, because it converts the control from an overhead into an insurance policy people remember buying.

Organisations that try to buy maturity with a framework purchase usually end up at level two with better templates. The framework is not the thing. The willingness to be governed is the thing.

Levels and dimensions are not the same measurement

Two things get called PMO maturity and they answer different questions.

Levels answer how far. They describe the overall state of the practice, and they are what most published maturity models measure. Dimensions answer where. They break the practice into components so that a single score does not hide a specific weakness, because a PMO can be strong on delivery discipline and have no strategic mandate at all, and one number will report that as average.

Both are useful and neither is sufficient alone. A level tells an executive whether the programme office is fit for the portfolio. A dimensional read tells them what to fund next. The Rydel Group PMO maturity self-assessment scores five dimensions, strategic alignment, governance, delivery, talent and capability, and tools and enablement, then places the result in one of three tiers. The tiers give the direction. The dimensions give the action.

Where the two disagree, trust the dimensions. An uneven profile is far more common than a uniformly mature or uniformly immature one, and the weakest dimension usually predicts the next failure better than the average does.