A project readout looks like a shorter consulting deck, and it's a different artefact with a different function. It belongs to a sponsor who commissioned the work and is now accountable for what happens next.
What are the five parts of a project readout?
The sequence is fixed: objective, findings, recommendation, next steps, then owners and dates. That order comes from governance, and persuasion has nothing to do with it.
So why not open with the answer, the way a consulting deck does? A consulting deck opens with its answer because the reader hasn't agreed to anything yet and attention has to be bought. A readout opens with the objective because the sponsor is checking, first, that you did the thing they asked for. Only once that's established do the findings matter. Put the recommendation on page 1 of a readout and a good sponsor will flip back looking for the objective. They can't evaluate a recommendation without knowing what question it answers.
Advertisement
| Section | Pages | What the sponsor is checking |
|---|---|---|
| Objective and scope | 1 | Did you answer the question I asked, and what did you exclude? |
| Findings | 3 to 5 | What is true, with the evidence, including anything inconvenient |
| Recommendation | 1 | What you think should happen, and what it is worth |
| Next steps | 1 | What happens in the next fortnight, concretely |
| Owners, dates and governance | 1 | Who has agreed to what, and when we check |
The Meridian Valley readout below runs to those proportions. Its objective page states that the sponsor asked whether the suite could absorb more cases without building rooms. It names the twelve weeks of block data used, and it states what was excluded, which is anything about surgeon-level performance. That exclusion line does more work than it looks like. It protects the analysis from a question it can't answer, and it tells the sponsor in advance that a certain kind of conversation isn't on the table.
The last section, owners and dates, is what separates a readout from a slideshow. It names a role for every action, a date for every action, and the governance point at which somebody checks. End a readout at the recommendation and you've handed the sponsor a document with no mechanism. The work then just doesn't get done.
The findings page: evidence before conclusion, and the finding you did not want
A findings page in a readout turns the consulting deck's habit upside down. The action title still states the finding. But the page underneath shows the evidence in a form the sponsor can interrogate, and the conclusion arrives at the bottom, where a consulting deck would have assumed it at the top. That's because a sponsor who's going to defend this work to their own board has to be able to reconstruct the reasoning. Having been persuaded by it isn't enough.
Worked example: the Meridian Valley findings page
Take the Meridian Valley suite. Eight rooms, five days, 480 minutes of prime time each, which comes to 320.0 hours a week of capacity. Service lines hold 270.0 hours of block, or 84.38% of prime time, and they use 198.8 of it. That's 73.63% block utilisation against a target of 80%, so the suite is 6.37 points short.
Every operating-room discussion reaches for turnover first, and turnover really is bad here. The mean over the six services is well above the 25-minute peer benchmark, and the excess totals 1,870 minutes a week, or 31.17 hours. But the released block, meaning hours a service holds and never uses, is 71.2 hours a week. That's 2.28 to 1 in favour of a problem nobody was talking about. Late first-case starts, the other usual suspect, contribute 7.05 hours a week, which is 6.45% of the 109.42-hour pool.
That's the finding. The page that carries it is a sorted bar of the three levers in hours a week with the target line drawn. It comes with one source line and one sentence naming what "released block" means operationally. Its conclusion, that a block-release rule outranks a turnover programme, comes at the bottom of the page and is repeated on the recommendation page. It stays out of the title, because the title states what was measured and the reader is being invited to check it.
What do you do with a finding that argues against you?
Now the harder part. Somewhere in a real analysis there's a finding that argues against your recommendation, and the temptation to leave it out is strong, sincere and wrong.
At Meridian Valley it's ENT. ENT has the best turnover in the suite at 26 minutes, one minute above the benchmark, and the lowest utilisation at 52.00%. For ENT, the turnover prize is 40 minutes a week, or 0.6667 hours, against 12.0 hours of released block. That's a ratio of 18.00 to 1. It looks like a counter-example, and it's actually the recommendation's strongest single case. A sponsor who finds it themselves after the meeting will wonder what else wasn't shown.
As a rule, a disconfirming finding stays in the deck, on its own page, with its own action title, and the page says what it costs the argument. Sometimes it costs nothing, as with ENT. Sometimes it costs a scope limit, which you state. Occasionally it costs the argument its conclusion, in which case you've learned something and the deck changes. What never happens is a quiet move to the appendix. The appendix is for evidence the body doesn't need. It's never a hiding place for evidence the body would rather not have.
Next steps somebody can start on Monday
A next step has four parts: an action stated as a verb, an owner named by role, a date, and a first move small enough to begin without a new approval. Anything missing one of the four is a wish.
Worked example: the Meridian Valley next steps
The Meridian Valley readout ends with five of them.
| Action | Owner (role) | Date | Annualised value |
|---|---|---|---|
| Block release rule at T-14 days | Director, Perioperative Services | Monday of week 1 | $5,367,747 |
| ENT block right-sized to the 80% rule | ENT Section Chief | week 3, next block cycle | $1,649,150 |
| Parallel induction pilot, two ortho rooms | Chair, Anaesthesiology | week 4 | $2,205,149 |
| First-case start huddle at 07:00 | OR Charge Nurse | Monday of week 1 | $1,329,372 |
| Weekly block scorecard published | OR Business Manager | Friday of week 2 | enabler |
Four things about that table are deliberate, and none of them is obvious the first time you see it.
Owners are roles and not people. The deck outlives the individual, and a named individual who wasn't consulted will reject the step in the room. The date is a date and not a quarter. You can check "week 3, next block cycle" on a Tuesday, while "Q2" is a way of agreeing to nothing. The value column is there because a next step without a value can't be prioritised against the other four. A sponsor with limited political capital will spend it on the first row ahead of the fourth.
Why can't you add up the value column?
The fourth thing is what separates a table you can trust from one that's only persuasive. These values aren't additive. Row 2 releases 8.750 hours of ENT block into the same pool that row 1 recovers from. At the 40% capture rate assumed in row 1, ENT's share is already worth $904,676 inside row 1. Adding the rows gives $10,551,417. Net of that single overlap, the figure is $9,646,741. Print the first number without the note and you've overstated the prize by 9.38%.
Disclose it on the page, in the supporting sentence, in one line: rows 1 and 2 overlap on ENT block; the combined figure net of double-count is $9.65m. A sponsor who finds a 9.38% overstatement themselves will discount every other number in the deck by an unknown amount, and they'd be right to.