After the first second, only one Meta placement changes as your ad plays

About 10.6 seconds into a vertical video ad in Meta's preview, Instagram Reels opens a card over the bottom of the frame that was not there a moment before. Every other main placement has settled by our first sample, at 1.2 seconds, and then stays exactly as it is.

We found that while placing type for a long motion video, one vertical master that had to work in Stories, in Reels and in feeds. The usual advice is a blanket band to keep clear at the top and the bottom of a 9:16 frame. It is written for the worst placement at the worst second, so it never says which placement is the strict one, and it says nothing about time. For a video that moves, time is the part that matters: if the thing covering the bottom only arrives at ten seconds, the first ten seconds can use that space.

So we measured it, per placement and per second. Everything here comes from our own test ad, not from a client account, and the data is in the downloads at the end of the post.

On a 1080 by 1920 file, the band from row 210 down to row 1,341, between x 60 and x 930, was clear in Instagram Stories, Facebook Stories, Instagram Reels and Facebook Reels at every second we measured. That is where the face, the headline, the product and the price go.

Below that band the frame is contested rather than forbidden. Until 10.6 seconds the four placements together leave the main column free down to row 1,411, and Facebook Reels is the one setting that limit. From 10.6 seconds Instagram Reels takes over, at 1,341. If you can leave Facebook Reels out of the placements, the first 10.6 seconds are free down to row 1,530.

The right edge is a separate matter. Both Reels placements put a column of buttons down the right-hand side, so anything the viewer must read stays left of x 930.

We created a paused draft ad in Riibon's own ad account and uploaded test clips at 9:16, 4:5 and 1:1. Each was a coordinate grid with a numbered line every 100 pixels and a running clock burned into the frame, and we added a clock-stamped copy of a real vertical clip.

The Advanced preview in Ads Manager plays every placement at once, each on its own looping clock. We zoomed on each pane every few tenths of a second, read where each element sat off the grid, and read when it appeared off the clock. Nothing was published and the campaign stayed paused.

Read the bars from the top. Instagram Stories and Facebook Stories show a top bar from the first frame and a button pill that was already up at our first sample, 1.2 seconds, and stayed at every sample after it. Facebook Reels is one fixed block across the lower part of the frame plus a column of buttons on the right, there from the first frame to the last.

Instagram Reels is the one with a story. A profile row and a caption line are there from the start. A slim button bar joins at about 2.5 seconds. At about 10.6 seconds that bar turns into a larger card with text, and the profile row moves up with it.

We could not tell whether the card opens on the video's own clock or on how long the viewer has been looking at the ad, because a looping preview cannot separate the two.

The explorer below draws the overlays at any second. Pick one placement or all four, drag the slider or press play, and the readout gives the lowest free row at that moment.

This is the lowest pixel row you can use before something covers it, in the main column from x 60 to x 930, keeping 25 pixels of margin above whatever comes next. The top edge, row 210, is the tallest top bar plus a wider margin.

At its worst, the bottom of a universal frame loses 579 rows to the overlays, about 30 percent of the height. The top loses far less than a blanket rule assumes: the tallest top bar we measured, in Instagram Reels, ends at row 170.

The two Stories placements leave the most room at the bottom. Instagram Stories is free down to row 1,595 once its pill is up, and Facebook Stories, which follows the same pattern with a swipe chevron above a larger pill, down to row 1,605.

Facebook Reels covers the lower part of the frame from row 1,436 with its profile row, caption, button bar and Ad tag, and its button column starts at row 1,330 on the right. None of it changes while the ad plays.

Audience Network, the placement in other companies' apps, puts a top bar over the video and a card with the caption and button across the bottom from row 1,430. When the video ends it is replaced by an end card. Facebook in-stream Reels and Ads on Facebook Reels showed a small sponsored card with a thumbnail and did not play the video in the preview, so for those the cover frame matters more than the timeline.

A 9:16 file shown in a feed is cropped. Facebook's feeds crop it to 4:5 from the middle, keeping rows 285 to 1,635, and Marketplace crops it to 1:1, keeping rows 420 to 1,500. Instagram feed draws the profile header over roughly the first 110 rows.

A 1:1 or 4:5 file is shown whole in feeds, with nothing drawn on top of the video. In Stories it sits in the upper part of the screen with the ad text and button underneath it. In Reels it is centred between black bars, and the caption and buttons fall on the bars and the lower edge of the video. Instagram Explore home did not accept our 1:1 file at all.

Instead of one safe zone we work in layers, because most of the bottom of the frame is not forbidden, it is contested, and a design can use contested space.

Always clear, rows 210 to 1,341 and left of x 930: the face, the headline, the product, the price, anything the viewer must read. Free until 10.6 seconds, rows 1,341 to 1,411: let graphics run on through it, but text there can be covered later. Covered often, rows 1,411 to 1,595: free in both Stories placements, free down to row 1,530 in Instagram Reels before its card opens, and covered by Facebook Reels throughout. Background only, the top 210 rows and everything below 1,595: colour and footage, nothing the viewer needs.

The data file in the downloads lists every element with its start, end and box, so a video generator can ask for the free rectangle at any second instead of reading a guide. That is how we use it.

Anything in Instagram Reels and the two Stories placements after about 22 seconds, our last sample of them; we assume the card and the pills stay to the end of the 30 second clip and beyond. Whether that card opens on video time or on time spent on the ad. The first second of Stories, because the button pill was already there at our first sample, at 1.2 seconds. The live apps, which a screen recording from the preview's share link would settle. And how an account's impressions split across placements, which is why every placement counts equally in the coverage map above.

The method is cheap, so it is worth repeating whenever Meta changes a placement: a grid clip with numbered lines every 100 pixels and a clock in the corner, a paused draft ad, the Advanced preview, and a note of which rows each overlay covers and when it arrives.