Board Deck Product Section: 7-Slide Skeleton

Ten minutes of board time, seven slides, zero feature parade. The board does not need to see the product. It needs the scoreboard, the bets, what you killed, the one risk that matters, and what you need decided. SCQA runs underneath: slides 1 and 2 are situation, slide 5 carries the complication, slides 3, 4, and 6 are the answer and the ask.

Time budget: 10 minutes presented, built to survive being read alone. Half your board reads the deck the night before and never hears you talk. Every slide must work without narration.

---

Slide 1: The scoreboard

The job: the five numbers that say whether the product is working, same five every quarter, with trend.

[Five numbers, large, each with: current value, last quarter, the trend arrow,
and one-line context only where the trend needs explaining.]

1. Direction metric: [The one number that says the product is getting better,
   not just bigger. Quality-of-outcome shaped, not vanity-shaped.]
2. Eval trend: [Composite or headline eval score on your production AI
   features, with direction. Boards in 2026 should see this number every
   quarter. If yours has never seen it, this slide is where that changes.]
3. Margin: [Gross margin or cost-per-outcome on the core workflow. The number
   that connects product decisions to the P&L.]
4. Adoption: [The usage number that predicts revenue: weekly active accounts
   on the core workflow, feature attach on the expansion driver. One, not six.]
5. Retention-relevant signal: [The product number that moves NRR: expansion
   trigger adoption, at-risk-account product signal, seat utilization.
   Whatever your CS and finance teams would agree predicts the renewal line.]

Speaker notes guidance: one sentence per number, prepared, only for numbers whose trend changed. Do not narrate flat numbers. The rule: same five numbers every quarter. A scoreboard that changes composition each board meeting is not a scoreboard, it is curation, and boards can smell it. Time: 2 minutes.

Slide 2: What changed since last board

The job: the delta, not the catalog. Three to five items maximum.

[Three to five bullets. Each: what changed, and the number or evidence that
proves it mattered. "Shipped [X]; [metric] moved from [A] to [B] in [n]
weeks." A ship with no number attached is not ready for this slide.]

[Include the bad news here, plainly, with what you are doing about it.
A product section with only good news teaches the board to discount all
of it. The credibility of slide 3 is purchased on this slide.]

Speaker notes guidance: for each item, the one question a board member is most likely to ask, with your two-sentence answer written out. Time: 1.5 minutes.

Slide 3: The bets table

The job: where product investment is concentrated, how each bet is doing, and how sure you are.

| Bet | Thesis (one line) | Status | Measure + target | Confidence | Calls it |
|-----|-------------------|--------|------------------|------------|----------|
| [Name] | [The causal claim] | [On track / Behind / Ahead / Pivoting] | [Number, baseline, target, date] | [High / Medium / Low, YOUR confidence it pays, not effort status] | [The date we decide] |

Rules: two to four rows, never more. Confidence is allowed to go down between quarters, a confidence column that only rises is fiction. Status "Behind" with a clear next step reads better to a board than suspicious unanimity of green. Speaker notes guidance: for any bet whose confidence dropped, the reason in two sentences and what would restore it. Time: 2.5 minutes. The center of the section.

Slide 4: The kill list and what it freed

The job: prove the strategy is a choice. What you stopped, what it was costing, where the capacity went.

[Two to four items. Each one line: what was killed, what it was consuming
(people, spend, attention), and where that capacity is now pointed.
"Killed [X]: freed [n] engineers and [$] of compute, redeployed to
[bet from slide 3]."]

Why this slide exists: boards see roadmaps everywhere. A kill list with redeployment math is rarer and more trusted than any roadmap, because it shows opportunity cost being managed. If the quarter had no kills, say so and say why, do not pad it. Time: 1 minute.

Slide 5: The one risk that matters

The job: one risk, treated seriously, with the premortem result. Not a risk register.

[The risk, in one plain sentence.]

[Premortem result: we assumed failure 12 months out and worked backwards.
The most likely path to failure is (one sentence). The earliest signal we
would see is (one sentence, with the number we are watching). What we are
doing about it now (one sentence).]

[What would change our posture: the threshold that turns this from
"watching" to "acting."]

Rules: one risk. A slide with six risks communicates that you have ranked nothing. Choosing the one risk is the work; the board learns more from your choice than from the list. Speaker notes guidance: keep the second-place risk in your notes for the inevitable "what else worries you" question. Time: 1.5 minutes.

Slide 6: The asks

The job: the decisions you need from the board. Decisions, not updates.

[One to three asks. Each: the decision, the date you need it by, and the
consequence of deciding versus waiting. "Approve [X] by [date]; each month
of delay costs [concrete consequence]."]

[Acceptable asks: capital allocation, a hire above a threshold, blessing a
kill with customer-facing consequences, a pricing posture, an M&A or
partnership posture, introductions with names attached.]

[Not asks: "support", "alignment", "feedback welcome." If you have no real
ask this quarter, the slide says: "No decisions needed this quarter. Next
likely ask: [X], expected [quarter]." That sentence builds more trust than
a manufactured ask.]

Time: 1.5 minutes, and protect them. If earlier slides run long, cut narration there, never here. The asks are why the section exists.

Slide 7: Appendix

The rules:

- Everything the board might ask for, nothing you plan to present: full metric
  definitions, cohort detail, the complete bets backlog, eval methodology,
  competitive teardown, the full risk list.
- Every appendix slide is referenced from a numbered footnote on slides 1-6.
  An unreferenced appendix slide is a slide you were afraid to cut.
- No new arguments. If a claim matters, it lives on slides 1-6. The appendix
  is evidence depth, not a second deck.
- It is allowed to be long. Boards respect a deep appendix they never open;
  it changes how they hear the six slides they did read.

---

The 10-minute budget, assembled

Slide 1  Scoreboard            2:00
Slide 2  What changed          1:30
Slide 3  Bets table            2:30
Slide 4  Kill list             1:00
Slide 5  The one risk          1:30
Slide 6  Asks                  1:30
                              10:00

Rehearse it once, out loud, with a timer. Whatever runs over, cut the narration, not the slide. And remember the half of the board that only reads: if a slide needs your voice to make sense, rewrite the slide.

QA before the board meeting

[ ] Same five scoreboard numbers as last quarter, same definitions.
[ ] Slide 2 contains at least one piece of bad news, handled plainly.
[ ] No bet table row without a "calls it" date.
[ ] Confidence column has moved at least once in two quarters.
[ ] One risk, not a register.
[ ] Every ask is a decision with a date.
[ ] Every appendix slide is footnoted from the body.
[ ] Pre-wired: the board member most likely to object to the biggest ask
    heard it before the meeting.