Read insights before you reframe a catalogue of throwaways

Every other format page tells you how to change shape. This one tells you which shape, from your own posts. get_content_insights breaks down recent social performance by platform and format. get_social_analytics and analyze_content sit nearby. None of that reformats a file. edit_video does, and that is where the credits go. Insights on Monday and a folder of unlocked generates reframed by Wednesday is a data-informed leak.

Pick the right format for each platform is the job. Reading insights is cheap next to an export. The export still has a free 480p preview. A brand-new account with nothing posted has nothing to analyze. Competitor accounts are out of scope.

Insights are not a generate

The tool's own description positions it as a read before you spend an afternoon reformatting a back catalogue. It is not a mandatory gate. It is also not a reason to reformat candidates. Averages over a small number of posts are noisy. Check the top-three list before treating a format's average as a verdict. Best posting hour can be absent if timestamps do not support it.

Then reframe keepers. 9:16 as a religion is how you crop a YouTube master you never needed to crop. Insights exist to stop that. They do not exist to bless a bin of misses.

The AI video editor is the export door. Insights do not export.

What has to be true

• You have enough of your own posts, through Versely's social integration, to mean anything.
• You have files that are allowed to freeze.
• You will derive children from those parents, not generate a new native ratio for every row in the report.

If the account is empty, pick a format from the brief, generate, lock, then start making history. Do not fake a recommendation.

Order

1. Read insights (optional, when the account has data).
2. Lock the masters you will actually reuse.
3. edit_video toward the winning format. Preview. Export.
4. Caption on the child.

estimate_cost on the export, not on the read.

FAQ

Can I use this to spy on a competitor's best ratio?

No. Your posts only. A competitor's 9:16 average is not in the payload.

Should I reformat the whole archive the day insights land?

Reframe the keepers that will get another life. Do not batch-export every unlocked experiment because a chart liked 9:16.

Is this a substitute for picture lock?

No. A format recommendation is a destination. Lock is which file is allowed to travel.

What if insights say 1:1 but the take is a vertical close-up?

Believe the take. Insights are a prior. A face that cannot survive 1:1 should stay 9:16, or you needed a different take — generated after you knew the destination, not cropped out of panic.