You have the data. Sales by month, ad spend by channel, orders by product. It is all in a spreadsheet, it is all correct, and looking at it produces exactly one feeling: there is probably something in here.
The gap is not the numbers. It is that finding a pattern requires knowing which pattern to look for, and pivot tables only answer the question you already thought to ask.
This prompt inverts that. You paste the data, state what you want to understand in plain language, and it comes back with the key conclusions, the trends and anomalies with actual numbers attached, and two or three recommendations. The anomalies part is often where the money is - the month that broke the trend usually has a reason, and the reason is usually actionable.
Run it in the AIMIKA chat in Text mode. You can paste the rows directly or attach the file.
You get the key conclusions in plain language, the trends and anomalies with the actual numbers attached, and two or three recommendations. The anomalies section earns its place: the month that broke the pattern almost always has a cause, and the cause is usually something you can act on.
One rule before you act on anything: check the numbers it quotes against your sheet. Arithmetic over long tables is a genuine weak spot for language models, and a confident wrong figure in a report is much worse than no report. Treat the output as a set of leads to verify, not a finished analysis.