How to Measure What Your Status Report Can't Show You

Your status report answers one question: What happened this week?

It does not answer the questions that actually stall programs:

• How long has this decision been open?
• How many forums did we hold before we closed it?
• Did we already decide this—and reopen it under a new name?

Those are decision latency fields. This essay is the procedure to capture them. Vocabulary lives on the Coordination Tax framework page; the worksheet is The Afternoon Test.

Step 1 — Pick the decisions that matter

Do not log everything. Log committed decisions—calls that unblock delivery, spend, or scope.

For one portfolio or program, start with five to ten rows:

1. A launch or release date call still open past the original target window.
2. A vendor or build-vs-buy fork that has survived three steering meetings unchanged.
3. A scope cut everyone agrees is needed but nobody will sign.
4. A dependency handoff where two teams are assuming different dates.
5. Any decision that was "resolved" last month and is back on this week's agenda.

If you cannot name five, ask your leads: What call are we circling that would unlock the critical path if we made it today? Write those names down. That list is your first log.

Do not log general risks, informational updates, or tasks with a clear executor and no decision gate. If the row does not require a named owner to choose between real options, it belongs in a RAID log—not in a latency worksheet.

Do log decisions that were "parked" with a softer label: monitor, socialize, align offline. Parking is often an open decision with better PR.

Step 2 — Capture four fields per row

For each decision, record:

Add an owner when you can name one human accountable for closure. If the owner column stays empty for two weeks, that is data too.

Decision latency is simple arithmetic: days from date raised to date decided. Open rows have no latency yet—they are your backlog of unpaid closure.

Worked example: A scope-cut decision raised on 1 June, closed on 18 June, with four alignment forums and no reopening → 17 days latency, forums = 4, re-decision = No. If the same scope cut reappears on 10 July as "phase-two descope," mark the original row re-decision = Yes and open a new row only if the substance changed.

Run the math in The Afternoon Test or on paper. The tool persists locally and exports a PDF for sponsor readouts.

Step 3 — Run the test on a fixed cadence

Name it the Afternoon Test because the signal shows up when the week is already bleeding—often Wednesday afternoon, when high performers realize Thursday is lost to rework and re-litigation.

Pick a cadence you can defend:

• Weekly for active transformation or launch portfolios.
• Biweekly for steady-state programs with lower decision churn.

Same day, same thirty minutes:

1. Update open rows (new forums, new owners, closures).
2. Add any new committed decisions from the week.
3. Mark re-decisions honestly—same substance, new meeting, new deck.
4. Read the summary: open count, average latency on closed rows, re-decision count.

Do not debate why the numbers look bad in the measurement block. Debate belongs in the closure block, after the log is current.

Consistency rule: Use the same definitions every week. If forum counts jump because someone started counting standups that were not about the decision, you lose comparability. Write your counting rules in the program name field of the worksheet once, then follow them.

Step 4 — Apply the Point of Sufficiency before the next forum

Before you schedule another alignment session for an open row, run the Point of Sufficiency check from the Coordination Tax term set:

Do we have the minimum information and alignment required for the accountable owner to decide—or are we gathering another round of input to delay the call?

If the owner is named and the facts are not in dispute, the next forum is Coordination Tax, not prudence. Close the row or escalate who can.

If facts are genuinely missing, name what is missing, who will fetch it, and the date it arrives—then hold forums until that date, not indefinitely.

This step is not philosophy. It is a gate on your calendar.

What to do with the readout

Bring three numbers to your sponsor or steering forum:

1. Open committed decisions — how many calls are blocking the critical path right now.
2. Average latency (closed rows) — how long closure actually takes in this program.
3. Re-decisions this period — how often you pay twice for the same call.

Pair the numbers with one sentence of context per open row: owner, blocker, next action date.

That readout replaces slides full of green indicators that describe motion without closure. It also gives you a baseline: if draft quality or AI throughput improves this quarter and latency does not, you are not more productive—you are producing more material around the same unpaid decisions.

When the pattern is portfolio-wide

A log fixes one program's visibility. If the same latency and re-decision pattern repeats across portfolios, measurement has done its job—you have evidence for a governance intervention.

For regulated AI rollouts, the next move is often a published boundary: Decision Rights Charter Sprint or the Charter Builder for a team-scale draft.

For execution drag without an AI angle, start with execution drag & governance and the Leadership Assessment to see whether deferral is personal, structural, or both.

The audit question

Open your calendar for the last two weeks. Pick one decision that consumed multiple forums.

Can you state—without opening a deck—date raised, alignment forum count, and whether it was a re-decision?

If not, you are flying on narrative, not measurement. Run the Afternoon Test once. The status report can stay green. Your log will tell the truth.