Lessons 67.4 and 79.5 asked for measurement without mythology. This appendix is the starter scorecard: one Markdown table plus an optional CSV, six columns, filled weekly in fifteen minutes. It tracks the complete path from exposure to cost — not the first number alone.

The six columns

• Exposure: how many qualified people encountered the message — impressions, sends delivered, aisle views. Reach, not vanity.
• Visit: how many arrived at the destination — page visits, listing opens, video viewers past the hook.
• First value: how many completed the smallest meaningful win — template saved, sample run, booking made, brief completed.
• Return: how many came back within the expected window — second open, second visit, saved-work revisit.
• Lead: how many raised a hand for follow-up — subscriber, waitlist signup, qualified booking, activated workspace.
• Cost: time plus spend that week — hours and cash, honestly logged.

One row per route per week. Three routes means three rows — the portfolio from Lesson 79.5, not every channel you have ever touched.

The starter files

# DISTRIBUTION-SCORECARD.md — [Project]

## Week of [date] — owner: ___
| Route | Exposure | Visit | First value | Return | Lead | Cost (time+spend) |
| --- | --- | --- | --- | --- | --- | --- |
| Compounding: ___ |  |  |  |  |  |  |
| Test: ___ |  |  |  |  |  |  |
| Direct: ___ |  |  |  |  |  |  |

## Learning question this week
- ___

## Vanity number we will not optimize yet
- ___ (e.g., total followers, raw impressions)

## Keep / change / stop note (one line per route)
- ___

week,route,exposure,visit,first_value,return,lead,hours,spend
2026-09-07,newsletter,2100,640,180,95,40,3,0
2026-09-07,marketplace-listing,380,120,14,5,6,4,0
2026-09-07,paid-test,5400,210,22,4,5,2,120

Filled example: one week, three rows

Week of 2026-09-07, Research Desk, owner: founder, fifteen minutes Friday:

Learning question: "Does the listing's setup step 2 lose people?" Vanity number ignored: total X followers. Notes: keep the note; change listing step 2 wording; stop nothing yet — test decision date is in three weeks.

Rules that keep it honest

• Fill every Friday before any new work. Missing weeks destroy the slope.
• Diagnose by empty stage (67.4): no exposure is a channel problem; exposure without visits is packaging; visits without first value is destination; value without return is fit.
• Log hours alongside spend. A "free" route costing eight hours a week is not free.
• Keep the CSV optional but identical — Markdown for humans, CSV when a tool must sort.

Verify fast

Cover the route names. Can you still tell which row is compounding, test, and direct from the numbers and costs alone? If not, the routes are not distinct enough — rename until the jobs differ.