Evidence of Value at the Budget Round

"$80M governed" is a denominator, not a value. It describes exposure, not effect, and a CFO will read it exactly the wrong way: the cost of oversight scales with spend, so the PMO is a tax on the portfolio. The dashboard has the same defect - on-time against a possibly re-baselined plan, budget variance against estimates the PMO may have helped write, and project counts are all activity and compliance measures. None of them tells the CFO whether decisions got better.

A value case should contain four things:

1. Decisions changed, with money attached. Projects stopped, re-scoped or resequenced before significant spend; dependencies caught before they cost a release; cost of delay avoided. Five concrete examples with numbers beat any portfolio-wide claim.
2. Outcomes against a baseline. Benefits realised versus the approved case, with named owners, and a stated comparison (before/after, or work that went through a given service versus work that did not).
3. Cost of the PMO against those effects, with assumptions exposed so the CFO can challenge them.
4. The counterfactual. What would be lost at a 30% cut. That is the decision actually on the table, and a value case that does not answer it is not a value case.

Fewer, verifiable, auditable claims; no framework language.

The 60% reporting effort is the target. Do a consumption audit across every recurring report: who reads it, what decision it drives, what happened the last time it was late. Retire or automate what fails the test, and quantify the FTE released - that becomes the funding for benefits work, so the change is self-financing rather than a request for more.

In parallel, address the benefits gap. Thirty percent of projects with no named benefit owner is a governance failure the PMO should name, not quietly absorb. Give the CFO the list, and propose a rule: no project passes the funding gate without a named benefit owner and a measurement plan. That costs the PMO nothing, improves the CFO's own reporting, and positions the PMO as protecting the CFO's investment rather than adding process.

Credibility with a CFO comes from speaking their frame: same currency, same baseline, assumptions exposed, and a commitment you can be judged against - "we will retire these five reports, redirect two FTE to benefits validation, and report realisation on the top 20 projects in two quarters". Do not propose a new framework or a new tool. Propose a smaller service portfolio with a price on it and a metric attached.

The trap is claiming outcomes the delivery teams and the business produce. A PMO that says "we delivered £12M of benefit" invites one question it cannot answer, and loses credibility permanently when it cannot.

Use three tiers, labelled honestly:

• Controlled by the PMO. Decision turnaround time, forecast accuracy at mid-point, intake cycle time, report consumption, share of projects with a validated benefit. These are defensible without any causal argument - you own them and can be held to them.
• Influenced, with a comparison. Projects that went through triage versus those that did not; dependency-flagged work versus unflagged; cost of delay on governed versus ungoverned change. Present these as associations, state the confounders (selection effect, project size, sponsor seniority), and do not add them into a total.
• Outcomes. Business results with a named owner; the PMO's contribution described, not claimed.

Verify by sampling: eight to ten projects, interviewed, with artefacts checked, rather than portfolio-wide assertion. Publish the assumptions and invite the CFO to attack them. A value case that survives scrutiny is worth more than one that inflates, and the budget decision happens once - the credibility lasts longer.