Reporting to a label without losing the argument

A monthly update to a label is not a data exercise, it is a negotiation with numbers in it. The manager who arrives with context and a recommendation gets a decision. The one who arrives with a screenshot gets questions.

Agree the definitions first

The most common way one of these goes wrong has nothing to do with the campaign. You say streams were up 9 percent, somebody else has a number showing 4 percent, and the next twenty minutes are about arithmetic rather than the project.

Both numbers are usually right. One is a trailing 28 days and the other a calendar month. One includes a platform the other excludes. One is corrected for reporting lag and the other is not.

Fix it once, in writing, before it matters. Which window, which platforms, which source, corrected or not. It is a two paragraph document, and it converts a recurring argument into a footnote.

Structure that survives disagreement

One paragraph up front: what happened and what you want. Not a summary of metrics. An argument. "The project grew through the month and the growth is concentrated in one market we have not been supporting. I want to move the remaining Q3 budget there."

Then the evidence, in the same order as the argument. Every number in the paragraph, in the order it was claimed.

Then the thing that did not work. Volunteered. This is the part that decides whether you get believed for the rest of the year.

Then what you cannot see. Which sources are not connected, and what that means for the report.

Then the ask, restated with a date. People forget the ask, so it appears twice.

Volunteer the bad number

If a campaign underperformed, say so first and with a figure attached. Two things follow.

You keep control of the interpretation. A number you raise comes with your explanation. The same number found by somebody else comes with theirs.

And you buy credibility for the good numbers. Anyone who has sat through enough of these updates knows which managers only bring good news, and discounts everything they say accordingly.

State what you cannot see

If a project has no advertising connection, the report says nothing in this report reflects paid activity because no ad account is connected. Silence on paid implies there was none.

This is the same discipline as an honest empty state in a dashboard, and it matters more here because the recipient is one step further from the data and cannot check.

What to include, and what to leave out

Include: streams gained in the window with the window named, monthly listeners and followers with their movement, the conversion ratio between them, geography changes worth acting on, campaign spend against the outcomes you can measure, live pacing per show if there are shows, and press count.

Leave out: total streams all time, impressions, playlist reach as a headline, anything cumulative, and any metric you have not used in a previous report. A new metric appearing in a month when the usual ones look weak is transparent to everyone in the room.

Keep the archive

Every monthly update, filed, with the numbers as they were reported at the time.

Two uses. It stops the slow revision of history where everyone remembers the last campaign going better than it did. And when a contract conversation arrives, six months of consistent, honest updates including the bad months is the most persuasive document you can produce, because it demonstrates something no single report can: that your numbers are reliable when they are unflattering.

The one sentence version

Say what happened, say what you want, show the evidence in that order, raise the bad news yourself, and be explicit about what you could not see. The credibility compounds, and credibility is what actually moves a budget.