This audit isn’t about whether your numbers are good. It’s about whether the way you read them, and the way you decide, can tell a real change from a normal week, and a good decision from a lucky one.

Open your weekly scorecard, the change log for your email and SMS flows, your last three launch or campaign plans, and your notes from the last few weekly meetings. Score each check 0 to 2: 0 if it failed or nobody can answer it, 1 if partly true, 2 if clean. “We usually talk about it” scores 0. The check is whether it’s written down.

If it isn’t written down before the result comes in, it didn’t happen.

The twelve checks

1. Weekly numbers have limits · 5 minLook at: Your weekly scorecard.
   Good: Revenue, orders, conversion rate and the main retention numbers each show a central line and upper and lower limits computed from their own recent history.
   Cost if wrong: Every wiggle gets a story and an owner, and the one change that matters looks like the rest.
   Read next: The Process Behavior Chart
2. Explanations wait for a signal · 3 minLook at: Notes from your last four weekly meetings.
   Good: Numbers inside their limits were noted and left alone. Explanations and actions were reserved for numbers outside them.
   Cost if wrong: The team learns causes that aren’t real and repeats them in next quarter’s plan.
   Read next: Most Movement Is Noise
3. Small denominators are labeled · 3 minLook at: Any rate on the scorecard built on fewer than about 1,000 people or 100 events a week: a segment’s click rate, a new SKU’s sales, a small cohort’s repeat rate.
   Good: Each shows its denominator, and nobody acts on a move that chance alone could produce.
   Cost if wrong: You kill good SKUs and promote lucky ones.
   Read next: Most Movement Is Noise
4. Open rate isn’t a decision metric · 2 minLook at: How email subject lines and sends are judged.
   Good: On clicks, orders or revenue per recipient, not opens.
   Cost if wrong: You optimize a number that Apple Mail inflates for reasons unrelated to your subject line.
   Read next: Most Movement Is Noise
5. Flows change on a plan or a signal · 4 minLook at: The change log for your five highest-revenue flows over the last six months.
   Good: Each change followed a signal on the chart, or was a planned test with a holdout. No flow was rewritten more than twice without one.
   Cost if wrong: Tampering adds variation, and nobody can say which version worked.
   Read next: Leave a Steady Process Alone
6. Winners are shrunk before they’re scaled · 3 minLook at: Your last three “winning” campaigns, creatives or SKUs that got more budget or inventory.
   Good: Someone estimated how much of the result would hold, and planned on that figure rather than the headline.
   Cost if wrong: You buy inventory and budget for a result that was partly luck.
   Read next: Every Winner Falls Back
7. Judgments of channels and people use two periods · 4 minLook at: How you last judged an agency, a channel or a team member on results.
   Good: You compared at least two separate periods, and the ranking held.
   Cost if wrong: You fire the unlucky and promote the lucky, then wonder why the new star fades.
   Read next: Skill, Luck and Persistence
8. Copied practices have a written reason · 3 minLook at: The last tactic you adopted because another brand did it.
   Good: Someone wrote down why it worked there, whether brands that tried it and failed exist, and why it should work here.
   Cost if wrong: You import a survivor’s story and none of its conditions.
   Read next: Other People’s Playbooks
9. Forecasts start from a reference class · 4 minLook at: Your last three launch forecasts: a product, a channel, a market.
   Good: Each began with what similar launches, yours or others’, actually did, and then adjusted for the specifics.
   Cost if wrong: Plans anchor on the best case, and inventory and hiring follow the plan.
   Read next: Start From the Base Rate
10. Live bets have kill criteria · 3 minLook at: Every initiative now running that costs real money or time.
    Good: Each has a written condition and a date at which it stops, decided before it started.
    Cost if wrong: Projects run on sunk cost and hope for a quarter longer than they should.
    Read next: Kill Criteria
11. A premortem before the last big launch · 2 minLook at: Your last major launch.
    Good: Before it went live, the team spent twenty minutes imagining it had failed and writing down why, and the plan changed because of it.
    Cost if wrong: The risks everyone privately saw stay private.
    Read next: The Premortem
12. A decision journal · 3 minLook at: Where big decisions are recorded.
    Good: Each has a dated entry written at the time: what you expected, how confident you were, and what would change your mind.
    Cost if wrong: Hindsight rewrites every decision, and you learn from outcomes instead of from reasoning.
    Read next: The Decision Journal

Score as you go; your band appears when all twelve are in.

Read your score

If checks 9 to 12 averaged lower than checks 1 to 8, start with part three. Reading numbers is something your team already practices every Monday. Writing things down before the result arrives is a new habit, and it’s the cheaper one to build.