Render Reels before your followers are active: use the Audience tab

Take the hour your followers are most active from the Audience view in Edits, then count backwards: the Reel should be rendered, checked and uploaded before that hour, not during it. Sume timeline renders are asynchronous jobs, so start them the evening before and let a webhook tell you when they finish.

NapoleonCat's roundup of Instagram's 2026 Edits changes lists audience demographics among the analytics; search coverage of the August Edits update says the Audience section shows where followers are, what they are interested in and when they are active (NapoleonCat, read 2026-10-07). Treat the active hour as a signal to test, not a promise.

A countdown that leaves room for a retry

Work back from the post time in the table. The slack is yours; Sume does not publish a render time guarantee, so leave room.

Why a webhook, and what to check

The default mode for timeline renders is async. Send mode: "webhook" with a public HTTPS webhook_url and Sume sends one terminal event, job.completed, job.failed or job.canceled. Verify the sume-v1 signature before you trust the payload, and keep a poll on GET /v1/jobs/:id/status as the fallback in case a delivery is missed.

The plan call checks the schema, the Sume-host URLs and the compile without creating a job, and returns billable_minutes. It cannot predict warnings about short sources that get padded or looped, so still watch the finished render.

What to prepare

Write down the hour you picked and the result. After four weeks you will know whether the active hour moved your views, and the log of job ids tells you which render each Reel came from.

Is the active hour worth chasing

Posting time is a weak lever compared with the Reel itself, and a single active hour is a noisy number. Use it as a tiebreaker: when two posting slots are equal for you, take the one when followers are online. Do not rearrange a week around it.

What the countdown does give you is slack. A render that finishes the evening before leaves time to find a typo in a caption, which is worth more than the hour.

Handling a failed render

Always leave room for one retry. If the job fails at the planned time, you want to learn that with hours to spare, not minutes. Use a fresh idempotency key for a retry after a failure, read the error code, and fix the input rather than resubmitting the same request.