White-Label Report Formatter

The CLEAN framework runs the build. Classify and group — sort the raw data into logical sections by business objective, not by platform. Language upgrade — replace platform jargon with business English ("CPC decreased 15%" becomes "cost per click dropped 15%, meaning each website visitor now costs less to acquire"). Extract key numbers — pull the three to five executive-summary metrics, with context on every one (vs last period, vs target, vs benchmark). Arrange in story order — lead with the answer, then the supporting detail, group by business objective rather than channel. Neutralise AI and tool fingerprints — strip the hedging openers, the platform watermarks, the export metadata, the AI rhythms that make a polished agency document feel mass-produced.

No one wants to read raw data. Every metric needs a sentence of context — what it means, where it moved from, what the reader should do about it. "CPC: £1.20" is a data point. "CPC: £1.20, down 15% from last month, meaning each visitor costs less to acquire and giving us room to scale the audience" is a report. The skill enforces the rule because it's the single biggest gap between an export and a deliverable.

The white-label scenario gets specific attention. If the report is going under another agency's brand, the language, voice, and detail level need to match what that agency would produce in-house — not the freelancer's house style under a different logo. The skill asks for the host brand's voice samples where possible and otherwise produces a deliberately neutral professional register that won't read as foreign in either direction. Platform fingerprints — column names from Google Analytics, ad platform terminology that signals the export source — get translated or removed before anything ships.

You get the formatted report with an executive summary that works as a standalone document for the 60% of readers who won't go further, a key metrics table with directional context and RAG status, narrative sections grouped by business objective, recommendations tied to the data, and an appendix for the detailed tables. The executive summary is written first because a CEO who reads only that should still walk away informed. Consistency is policed throughout — same date formats, same decimal places, same terminology — because inconsistency is the strongest signal that the document was assembled rather than authored. When data sources conflict, the skill flags the discrepancy openly rather than hiding it.