# Capture Report — design decisions (parked 2026-08-08, resume before first client month-end)

The monthly retention artifact promised on every flyer ("End-Of-Month Review Free — the Capture Report"). Nothing built yet; these are the agreed design principles from the bail-flyer session.

## Agreed

1. **One page, plain email, no portal** — these are no-SSL shops; a login link goes to die.
2. **HIS close rate, never industry stats.** Month one / the 25-call tuning session captures "of the intakes we handed you, how many wrote?" — HIS number runs the worth-math forever. Unimpeachable because it's his.
3. **The junk count is a feature line, not a footnote.** "47 telemarketers filtered — you never heard about a single one" itemizes invisible value. (Depends on the build-side junk filter: handoff text fires only when name + callback captured — the ⚠️ DO-NOT-FORGET build item in FLYER-REBUILD-PUNCHLIST.md item 6.)
4. **Per-call ledger** — every call one line (date/time · name · duration · intake|junk) + transcript link. He can audit the guarantee count himself.
5. **Never skip a month.** Zero-intake months still send ("your line was quiet; here's the 3 junk calls it ate"). Silence from us reads as silence from the system.

## Structure sketch

THE HEADLINE NUMBER (unreachable N times, answered all N) → THE LEDGER (real intakes ✅ = guarantee count · junk filtered 🔇) → WHAT IT WAS WORTH (N intakes × his close rate) → THE CALLS (ledger) → TREND (vs last month).

## Open questions (Rob to decide)

- Plain email vs PDF attachment
- Week-one "first catches" text after go-live (early proof while novelty's high) — yes/no
- Guarantee shown as running total ("14 of 30 by day 22" — scoreboard psychology) vs final month-end number