Drafting a monthly performance report from an analytics export
Monthly reporting to a client or a boss means pulling numbers out of each platform's native analytics and turning them into a narrative that says what worked and what to change. From ideation to reporting, AI gives social teams back time lost to manual workflows (a theme of the 2025 Sprout Social Index). The discipline is feeding it only real, aggregated numbers and never letting it guess at causes it can't see.
You are a social media analyst. Draft a monthly performance report from the data I paste. This period's metrics (aggregated numbers only — the only figures you may use): {{metrics}} Prior period for comparison: {{prior_period}} Goals we set for this period: {{goals}} Produce: - An executive summary in five plain sentences. - A table of key metrics with the month-over-month change and percent change. - Three things that worked and why, each tied to a specific post or format. - Two things to change next month. - Progress against each stated goal. Rules: - Use only the numbers I provided. Do not estimate, invent benchmarks or industry averages, or round beyond one decimal place. If a metric is missing, write "not provided." - Do not claim a cause you cannot see in the data. Frame any explanation of why a number moved as a hypothesis to test, not a fact. - Plain language, no jargon, written for a client who is not a social media expert.
Fill in your details and the prompt updates live — then copy.
Executive summary. September beat all three goals. Reach grew 16.1% month over month to 412,000 and engagement rate rose from 3.6% to 4.1%, clearing the 4% target. Link clicks reached 3,900 against a 3,500 goal, and the account netted 1,240 new followers. The Sept 8 schedule carousel was the standout at 22,000 reach. | Metric | This period | Prior | Change | % change | | Reach | 412,000 | 355,000 | +57,000 | +16.1% | | Engagement rate | 4.1% | 3.6% | +0.5 pts | — | What worked: the schedule carousel outperformed static posts (hypothesis to test — carousels may be favored this month; confirm before shifting the plan).
The full workflow
- Export aggregated metrics from each platform — totals and rates only, no follower-level data
- Paste this period and the prior period so the tool can compute real changes
- Run the prompt, then recompute the percent changes yourself to catch any arithmetic errors
- Replace any AI-stated cause with your own read of why numbers moved, or mark it as a hypothesis
- Add screenshots of the top posts and send to the client with your own recommendation on top
Watch out for
AI states confident causes for numbers it can't actually explain — attributing a reach spike to a tactic when a platform algorithm change or a single viral post was the real driver. Treat every why as a hypothesis and verify it before you change strategy on it.
Do not paste follower-level or customer-identifying data (names, handles tied to complaints, DM contents, purchase records) into consumer AI tools. Report with aggregated counts and rates, and keep identifiable data in your own analytics platform.
Where this comes from
Every use case on this site is grounded in real reports from working social media managers — not invented by us.