Display Banner Sets Designed Size by Size, Not Scaled

A display set is one argument compressed to different degrees, not one design scaled onto eleven canvases. A 320x50 holds four words and a mark. A 970x250 holds a sentence, a photograph with somewhere to sit, and a price. If those two files are the same artwork at different percentages, one of them is wrong, and it is almost always the small one. Where the work starts: Five sizes carry most of the delivery across the Google Display Network, Xandr and the Dutch publisher houses: 300x250, 336x280, 300x600, 728x90 and 320x50. Those five are designed as originals, with separate crops, separate copy lengths and separate decisions about whether the logo sits in a corner or owns the final frame. The remainder of the IAB portfolio is derived from whichever original it most resembles, which goes quickly precisely because the hard choices were made five times, not eleven. Copy is part of this, not an extra: A headline that survives at 320x50 is a different sentence from the one in the campaign deck. Asking a designer to fit the deck sentence into fifty pixels of height is how a set ends up carrying 9pt type nobody reads. Copy is signed off at every size before a single layout is refined. Adaptation, and its limit: If a key visual exists from a brand campaign I will adapt it, and adaptation means I can recrop, re-set the type, and remove elements the small sizes cannot carry. Where the agreement with your brand team is that the visual arrives locked and must emerge untouched at eleven aspect ratios, that is unit-priced production work and there are studios who do it properly. I am not one of them, and taking that brief would cost you more than it costs them. Disclaimers: For 'from EUR X' travel pricing, and for anything adjacent to mortgage or investment claims on property, the disclaimer is placed first and the composition is built around it. Retrofitting two lines of required disclosure into a finished 300x250 wrecks the layout and delays the flight. Send the legal text at briefing, in draft if that is all that exists. Delivery: A layered Figma file with the five originals as components and the derivations as variants, exports at 1x and 2x, weights under each network's static limit, filenames matching your ad server taxonomy so reporting groups without hand-tagging, and a size-by-size note of what was cut and why. Static artwork only; animation is a separate piece of work with a separate budget. Not included: HTML5 or animated versions of the set. Media planning, beyond telling you which sizes I think the buy needs. Photography and retouching, since catalogue shots with hard shadows show badly at 970x250 and no layout hides that. Print, out-of-home and in-store adaptation. Translation into languages beyond English and Dutch. Who this is not for: Buyers who want the full portfolio because it exists rather than because it is on the plan. I will ask which sizes the buy actually contains and quote for those, which usually reduces the invoice. Also not for teams whose offer is still moving: a banner set is where a campaign's argument is tested at its shortest possible length, and an offer that changes after briefing means the set is drawn twice at your cost. Second markets: Dutch and English versions are designed together rather than translated afterwards, because line lengths differ by roughly a fifth between them and that difference decides layouts, not copywriters.