What is Safe zone?

A safe zone is the region of an ad's frame that will still be visible and legible after the platform has done what it does to it: overlaid its own interface, cropped to a placement's aspect ratio, or both.

The frame an advertiser uploads and the frame a person sees are not the same. On a full-screen vertical placement, the platform puts a profile name and caption over the top of the image and a call-to-action button, progress indicators and interaction controls over the bottom. On in-feed placements the frame is trimmed to fit. On some surfaces both happen.

Whatever sat in the covered or trimmed regions is gone, and nothing in the account says so.

There is no single safe zone, and the reasons are structural rather than arbitrary.

Aspect ratio. A placement renders at a particular shape. An asset of a different shape is cropped to fit, and the crop takes the edges.

Interface density. A full-screen placement has more overlay than a small in-feed unit, because it has more interaction to offer.

Surface conventions. Different platforms put their controls in different places, and the same platform puts them in different places on different surfaces.

Redesigns. Interfaces change. A safe zone measured against last year's layout is a measurement of last year's layout, which is the main reason memorising specific percentages is a poor substitute for checking.

The practical consequence is that an account buying several placements is buying several different frames, and an asset composed for one of them is compromised on the others.

The failure mode is quiet, which is what makes it expensive.

A headline placed near the bottom of a vertical video sits under the call-to-action button. A price in the top corner disappears behind the caption. A subject's head is cropped out of a portrait render of a landscape photograph. A logo is half covered. A closing frame with the offer written across it is trimmed to a fragment.

None of this is reported. The ad serves normally, the impressions are counted, the spend is real, and the only symptom is performance worse than the creative deserved. An account diagnosing that will look at the audience, the bid, the offer and the landing page, and will not look at the frame.

Keep type and essential subject matter well inside the middle. Generous margins cost composition and save legibility, and the trade is almost always worth taking.

Compose for each shape. Vertical, square and landscape versions built for their shapes, not one asset cropped three ways. This is a production cost and it is the only reliable answer.

Do not put anything load-bearing at an edge. The edges are what the crop takes and what the interface covers.

Assume the closing frame will be interrupted. On placements where a person can swipe away, the end of a video is the least reliable place to put the message.

Check the preview. Every major platform offers a placement-by-placement preview before launch. Looking at each one takes a few minutes and catches the majority of these problems, and it is skipped constantly.

Platforms will generate placement variants from a supplied asset, resizing and cropping to fit surfaces the asset was not made for.

It is worth understanding what this is for. It prevents the worst outcome — an asset simply not serving on a placement, or serving letterboxed with bars — and it does not produce a good result, because it is working from a composition designed for a different frame. The crop it chooses is a guess about what matters in the image.

Where an account cannot supply every shape, automatic adaptation is better than nothing. Where it can, supplying properly composed assets is the difference between an ad that works on every surface and one that works on one and survives on the rest.

Safe zones are a specific instance of a general fact about modern advertising: the advertiser controls the components and the platform controls the presentation.

That is true of assembled text formats, of dynamic creative, of placement allocation, and of the frame itself. The response is the same in every case — supply material that works under any arrangement the platform might choose, rather than material that only works under the arrangement you had in mind.

For creative, that means margins, legibility at small sizes, subjects clear of the edges, messages that do not depend on the last frame, and a check in the preview before anything runs.

The teams that never have this problem solved it once, upstream, rather than checking each asset.

A template per placement. A working file with the covered regions marked, so that anything placed outside them is visibly wrong while the asset is being made rather than after it has run.

A shoot framed loose. Footage captured with room around the subject can be recomposed for several shapes; footage framed tight for one shape cannot.

Type on its own layer. Text that can be moved without re-rendering the image makes producing several shapes a matter of minutes rather than a new export each time.

A review step that looks at previews, not at the uploaded file. The uploaded file always looks correct; the preview is where the interface appears.

None of this is sophisticated and all of it is skipped under time pressure, which is why the same failures recur in accounts that know perfectly well what a safe zone is.

Three situations produce most of the damage.

Repurposed organic content. A post made for one surface, promoted on several. The composition was right for where it was made and is wrong everywhere else.

Landscape libraries. Brands with a stock of wide imagery pushing it into vertical placements, where the crop is severe and takes the subject with it.

Offer-led creative. Where the message is a price or a deadline rendered as type, the cost of that type being covered is the whole ad rather than part of it.

In each case the remedy is the same and the reason it is skipped is the same: producing for the frame costs something, and the failure it prevents is invisible in every report the account has.