As of 13 August 2026, AI can analyse your social media performance.

What to actually do

1. Hand it to a person

   The route this page recommends

   A person who owns the outcome does this end to end, worth it when the failure is dear.

2. Use a tool built for this

   Second choice

3. Do it yourself

   The distant third

   A chat interface, power-user skill, and roughly 30 minutes until you can act on the result.

   How to actually do it

   1. Open the analytics area for each relevant social platform and export the same reporting period, including post-level metrics, audience figures and paid campaign results where relevant.
   2. Put the exports into one clearly labelled spreadsheet, keep the original platform names and metric definitions, and add columns for platform, post format, topic, campaign and objective where those details are missing.
   3. Write down the business goal for the analysis, such as reach, engagement, website visits, leads or sales, and note any campaigns, posting changes or unusual events during the period.
   4. Upload the spreadsheet and the goal and context notes to an AI chat, then paste the supplied prompt beneath them.
   5. Compare the AI's totals and engagement-rate calculations with the platform dashboards and the source spreadsheet, correcting any mismatched definitions or missing rows.
   6. Ask a colleague who understands the account to challenge the explanations and proposed tests, then mark which findings are observations, hypotheses or unsupported claims.
   7. Turn the agreed tests into a tracking sheet with the chosen metric, baseline, content change and review date, and use the next platform export to compare the results.

   Prompt

   Analyse the attached social media performance data for [business or account] over [reporting period]. The data covers [platforms] and includes, where available, post date, format, topic, reach, impressions, views, likes, comments, shares, saves, clicks, follows, watch time, spend and conversions. My main goal is [goal], and my audience is [audience].
   
   First, check the column names, missing values, duplicate rows and inconsistent definitions. State any limitations before drawing conclusions. Compare performance by platform, content format, topic and campaign where the data supports it. Calculate engagement rates only when the required fields are present, show the formula used, and do not invent missing figures. Identify the strongest and weakest posts using clearly stated criteria, then separate direct observations from possible explanations.
   
   Produce:
   1. A short executive summary.
   2. A table of the key metrics and comparisons.
   3. Three evidence-based findings with the source rows or fields supporting each one.
   4. Three practical tests for future content, each with a hypothesis, audience, content change, success metric and review method.
   5. A list of claims that I must verify in the platform analytics or campaign records.
   
   Do not claim that one post or tactic caused an outcome unless the data supports that conclusion. Do not recommend spending more money without explaining the evidence and the uncertainty. Use plain British English.

   Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.