Reel analysis prompt
What worked? What should you try next?
Use your actual Instagram insights to make the next creative decision. This prompt helps an AI assistant separate the numbers, the interpretation and the next test.
Bring the
right evidence.
- Open the Reel’s insights in your professional Instagram account. Capture the visible metrics and, if available, the retention graph showing how viewing changes through the video.
- Add the video length, script and what you wanted viewers to do. A screenshot alone cannot explain the creative.
- Include comparable posts and any attributable contacts, bookings or purchases. Keep each platform and time period separate.
Remove names, email addresses and private customer details before uploading screenshots to an AI assistant. You are sharing with the provider you choose, not submitting those screenshots to Hanami.
Ready to use
The prompt
Paste this into your preferred AI assistant, fill in the inputs, and review its answer before using it.
Help me understand the results of an Instagram Reel and decide what to test next. INPUTS Business and product: Who the video was for: What we wanted viewers to do: Video length and posting date: Video script or description: Insights screenshots or typed metrics: Comparable recent videos, if available: Leads, bookings or purchases we can attribute to this video, if available: 1. Read the evidence. List each visible metric with its exact label, value, platform and time period. Keep Instagram and Facebook separate. Distinguish views, accounts reached, video-attributed follows, account follower changes, comments, shares, saves and contact records. If a value or graph cannot be read, write “unknown”. Do not fill gaps from assumptions. 2. Explain the result. Compare with relevant recent videos from this account when those are supplied. Explain what the reach, viewing time, retention graph and responses suggest. Define unfamiliar terms. Calculate rates only when the appropriate numerator and denominator are both available, and show the calculation. Do not apply universal pass/fail thresholds or invent an account benchmark. 3. Find the most useful improvement. Identify the strongest supported opportunity. Separate what the data shows from your interpretation. A retention drop does not prove why someone left: connect it to the script or footage only when that evidence is supplied. If the result is strong, identify what is worth repeating rather than manufacturing a problem. 4. Recommend the next tests. Give three specific changes to test, ranked by usefulness. Where evidence permits, refer to a moment in the video. Explain which outcome each test aims to improve: attention, understanding, contact capture or a buying action. Change one important variable at a time where practical. OUTPUT A short result summary. A table of the actual metrics, calculations and unknowns. The strongest opportunity, with evidence and uncertainty. Three practical tests. The missing evidence needed to judge business impact. Never treat views, followers or platform contacts as revenue. Do not claim an algorithmic cause, guaranteed distribution or sales lift from one screenshot.
You can also select and copy the text above.
Use it to make
a decision.
A useful answer tells you what happened, what remains uncertain and what to test. It should not turn every metric into a claim about the algorithm.
A video with fewer views can still bring better customers. Review creative performance alongside the action your business wanted: enquiries, product purchases, bookings or something else you can measure.
If the assistant cannot read a number, supply it yourself from the original screen. Never let it guess.
Put it to work
Good content is
only the beginning.
See how we connect videos, automated messages and sales. Then tell us what you’re selling.