The report

One report a month, on what each channel actually caused

Your channels each credit themselves, and added up they claim more revenue than your store took. This settles it: a number per channel, the range it sits in, and where the next dollar earns. The year so far is below, month by month, exactly as a paying brand receives it.

Synthetic data

Real output from the engine, run on a made-up brand. The numbers are invented; the reasoning is exactly what a client receives.

Open the full report
6things in every report
Every channeljudged on its own evidence, not given a slice of one total
The 3rdof every month, once the month holds still
Year so farin the first report, month by month

Six things, every month

Every channel you run gets its own number, the reasoning that produced it, and an instruction for the next dollar.

Judged on its own evidence, not handed a slice of one total. The third one is where you check the reasoning instead of trusting it: every conclusion arrives with the ordinary explanations that were set aside, and why.

  1. 01

    The number, and its range

    A width, not a point. When the range is wide the report says so instead of picking a tidier number.

  2. 02

    Made the demand, or caught it

    On a dashboard these look identical. For your budget they are opposite.

  3. 03

    What was ruled out, and why

    A conclusion without its rejected alternatives is an assertion. This is where you check the reasoning instead of trusting it.

  4. 04

    What caps the confidence

    Every report names its own ceiling. You always know the ground the number stands on.

  5. 05

    What would change the verdict

    The signal that would move it next month. Proof the conclusion is not dogma, and a list of what to watch.

  6. 06

    The recommendation, and its guardrail

    Where the next dollar goes, per channel. A channel that returns well and still loses money on every new customer it brings gets a fix, not a budget increase.

A strong return can still be the wrong move

A channel can return well and still lose you money on every new customer it brings. When that is true, the report tells you to fix it rather than to scale it.

The checks that catch it are worked out from your own ledger rather than asked for, and any one of them can stop a recommendation the efficiency case would have waved through.

Before any recommendation to spend more
Scale Metathe efficiency case is strong
Platform ROASdo the platforms’ own claimed sales cover the spendClears
Contribution floordoes the revenue cover the spendClears
Payback on a new customerdoes the first order cover what it cost to win itFails
Not recommendeda strong return means nothing if every new customer costs more than they spend
Released to youNot reached
Each one is worked out from your own ledger, and each can stop a recommendation on its own. Every check behind a recommendation is named on the report.

On day one, and the 3rd of every month

Connect your store and the first report covers the year so far, month by month. After that a new one lands on the 3rd.

Not the 1st, and the two days are the difference between a report and a guess. A month does not stop changing when it ends: every ad platform keeps rewriting its recent days while attribution settles, and refunds keep landing in your own ledger. We wait for the month to hold still.

Nothing to operate to get it. It arrives written, in language a CFO can challenge you on, and reading it is the whole job.

Read one before you decide anything.

Every channel, month by month: what each one claimed, and what the evidence actually supports.

Get your first report