River
Y CombinatorBacked by Y Combinator
FREE TEMPLATE

Leadership Decision Log Template

Two documents and three sheets, including a Reopened Decision Analysis that checks whether a recorded dissent predicts fewer relitigated decisions than one logged as unanimous.

Free download  ·  No account needed

Reopened Decision Analysis

No-dissent decisions reopen 4.67 times as often

Halberd Instruments, eight-person leadership team. Twenty-six weeks of Monday decisions, 9 February to 6 July 2026.

GroupTopicsReopenedReopen rate
Dissent recorded at original decision7114.3%
Logged as unanimous ("none recorded")12866.7%
All nineteen topics19947.4%

Of the eleven reopens across those nine topics, five named a new fact, a crossed threshold or a failed first action. The other six named none of the three, and every one of those six traces back to a topic originally logged as unanimous. Not one of the seven dissent-carrying topics has a reopen that names nothing.

A decision log template asks for the same three things every time: what was decided, who owns it, and the rationale for the winning option. Page-one results for a leadership team's own decision log, from a four-field template to a full leadership tracker, all stop there. None asks who disagreed at the time, on the assumption that a sound rationale already covers it. This pack tests that assumption against a real register instead of taking it on faith.

Halberd Instruments, a fictional instrumentation manufacturer, ran its eight-person leadership team's Monday meeting through a Decision Register for twenty-six weeks, 9 February to 6 July 2026. Every one of its nineteen decisions carries a dissent field, filled in either with a name or with "none recorded," never left blank, because a blank cell only means the question was never asked. Seven decisions carried a named dissent. Twelve were logged as unanimous. Checked against which decisions actually got reopened later, the two groups did not behave the same way at all.

"No dissent recorded" decisions reopened 66.7 percent of the time, eight of twelve topics. Dissent-recorded decisions reopened 14.3 percent of the time, one of seven. Of the eleven reopens, five named a genuine new fact, a crossed threshold, or a failed first action, and six named nothing at all. Every one of those six traces back to a topic originally logged as unanimous, and none traces back to one with a dissent on file. The transcripts behind the register are the same ones meeting notes and action extraction already turns into owned commitments.

Thirty decision rows, a dissent field on every one, and the analysis it feeds

The Decision Register, the Follow-through Tracker, and the Reopened Decision Analysis sheets from the pack.

Decision Register

Halberd Instruments, Monday leadership meeting, 9 Feb to 6 Jul 2026. Dissent is a name or an actively written "none recorded," never a blank cell.

IDDateTopicDissentReopen of
DL-00109 FebFontana forecast ownershipNone recorded
DL-00316 FebVanterre sourcingCallista Verhoeven
DL-00702 MarPlatform rebuild budgetMarguerite Sable
DL-01606 AprFontana forecast ownership, reopenn/aDL-001
DL-02311 MayVanterre sourcing, reopenn/aDL-003
DL-02715 JunFontana forecast ownership, reopenn/aDL-001
19 original decisions + 11 reopens = 30 rows. 7 topics carry a named dissent, 12 carry "none recorded."

Fontana's account-ownership split is the register's most-relitigated topic: logged with no dissent in February, reopened twice with no new fact named either time. Vanterre's sourcing call carried a dissent from the start and reopened once, in May, when a named fact (two missed delivery windows) actually changed.

Follow-through Tracker

One row per decision's first committed action, checked against its own target date.

IDFirst actionOwnerStatusDays late
DL-001Publish the 30-day handoff checklistDuncan MhlabaLate14
DL-004Update org chart, notify cohortDuncan MhlabaNo action stated
DL-006Notify secondary carrier of wind-downCallista VerhoevenNot done
DL-012Send MFA enrollment deadlinePerpetua NwosuLate5
DL-020Notify supervisors of new thresholdCallista VerhoevenLate4
30 first actions tracked: 19 on time, 7 late, 3 not done, 1 with no first action stated at all.

A decision with no stated first action is not assumed done. DL-004 carries no date and no owner-side confirmation, so the tracker flags the gap rather than crediting a follow-through that was never actually checked.

Reopened Decision Analysis

Every reopened topic, grouped by whether its original decision carried a dissent, checked against the Reopening Criteria.

TopicOriginal dissentReopensMet a stated criterion
Fontana-forecast-ownershipNone recorded2No, both times
Vanterre-sourcingCallista Verhoeven1Yes
ECR-signoff-orderNone recorded1No
freight-carrierNone recorded1Yes
field-tech-reportingNone recorded2No, both times
vp-equity-signoffNone recorded1Yes
11 reopens total: 5 meet a stated criterion, 6 do not. All 6 that name nothing trace to a topic originally logged with no dissent. None of the 7 dissent-carrying topics does.

Reading the two columns together is the finding: a dissent on file is not what keeps a decision from being questioned again, since Vanterre reopened too. What a dissent on file predicts is that the reopen, if one happens, will name an actual reason.

What is in the pack

01

Decision Register

One row per decision, carrying a dissent field that is always actively filled in: a name, or "none recorded," never blank. Nineteen topics and thirty rows counting reopens, made in the same Monday meeting an executive calendar audit would price by the hour.

02

Follow-through Tracker

The first action every decision commits to, its target date, and whether it happened on schedule: nineteen on time, seven late, three never done, and one with no first action stated at all.

03

Reopened Decision Analysis

Groups every topic by whether its original decision carried a dissent, then reports the reopen rate for each group side by side: 14.3 percent against 66.7 percent on the worked example built into the pack.

04

Decision Log Standard

States what belongs on the register and the line between a blank cell and one that actively records agreement, which is the rule the rest of the pack is built on. The same team also meets twice a year for a longer session, where a proposed offsite topic gets checked against an older register instead of a fresh judgment call.

05

Reopening Criteria

The three facts that make a reopen legitimate rather than a relitigation: a new fact, a crossed threshold, or a failed first action. A returning topic missing all three is named as a relitigation in the room rather than logged as though something changed. A related discipline governs a different register: inbound request triage uses a fixed rule instead of a fresh judgment call to decide whether Thea's executive assistant escalates a request, delegates it with a heads-up, or auto-delegates it.

How it works

  1. 1

    Send transcripts, notes, or any decisions already logged

    Whatever exists: meeting transcripts, informal notes, or a spreadsheet nobody has updated in months. Partial history is fine, and it is usually where the missing dissents are hiding.

  2. 2

    Every decision gets a topic tag and a dissent field

    Filled in with a name or with "none recorded," never left blank, because a blank cell and an actively written "none recorded" are different claims, and only one of them is evidence of agreement.

  3. 3

    A returning topic gets checked against the reopening criteria

    A new fact, a crossed threshold, or a failed first action. Where none of the three is named, the return is flagged as a relitigation instead of logged as though something changed.

  4. 4

    See the reopen rate, then carry the register forward

    The Reopened Decision Analysis reports the reopen rate for dissent-recorded decisions against unanimous ones, on your own register. Keep logging with the same dissent field every meeting after that.

Frequently asked questions

Why record who disagreed if the decision is going ahead anyway?

Because the decision proceeding and the disagreement being real are not in conflict. Halberd's CFO, Marguerite Sable, dissented when the platform rebuild programme got its own budget line, preferring it stay inside core engineering's budget. The programme got the line anyway. Her objection is on record; the decision stands unchanged.

How is this different from a decision-rights document?

A decision-rights document, such as one tested against what an escalation actually changed last quarter, settles who gets to decide. This register assumes that question is already answered and tracks what happens after: whether the decision that got made stays made, or quietly comes back.

What counts as a legitimate reason to reopen a decision?

One of three things, named explicitly on the reopen's own row: a specific new fact, a threshold now crossed, or the decision's own first action failing, as tracked on the Follow-through Tracker. Formal governance treats reopening the same way: a motion to reconsider a vote is only timely the same day or the next, before the room has moved on.

Why write "none recorded" instead of just leaving the field blank?

A blank field cannot be told apart from a question nobody thought to ask. People often withhold disagreement out of fear of the room's reaction, not because they agree: Edmondson's original study of team psychological safety found reluctance to speak up tracks perceived interpersonal risk, not the absence of a concern. Writing "none recorded" forces the question out loud, every time.

What happens to a decision once it is logged?

The Follow-through Tracker records a decision's first action and whether it happened on schedule; this pack stops there. A decision that reaches the board carries forward into the reconciling register the board packet keeps under its own arithmetic identity every meeting.

Does recording a dissent mean whoever disagreed was right?

No, and Vanterre Group is the counter-example on the register. Callista Verhoeven dissented in February; the single-source decision proceeded anyway, and it was still the right call in February. It only reopened in May because Vanterre missed two consecutive delivery windows, a new fact, not because the original dissent was vindicated by hindsight.

What format are the downloaded files, and is it free?

Two Word documents and three CSV sheets in one zip, free and with no account. The documents open in Word, Pages and Google Docs, the sheets in Excel, Numbers and Google Sheets. Edit with AI installs the same pack as a private Space and checks the reopen rate against your own register.

Find out whether your own decisions are staying decided

Send your last few months of leadership meeting notes or transcripts, plus whatever decision log already exists. The first thing back is which of your decisions were logged with a dissent and which as unanimous, before any reopen rate gets computed.

Edit with AI