A national retailer runs on the promotional calendar. The creative job isn't to make a sale look beautiful, it's to make the offer land before the thumb moves.
I sit on the in-house team at Furniture Village, working to briefs from marketing and trading. The rhythm is relentless and the dates are fixed: Winter Sale, Spring Clearance, Early Bird Offers, Home Event, Bank Holiday, Black Friday. Every campaign needs animated creative across several placements, mostly Meta and TikTok, mostly vertical, always watched with the sound off. The work below is a selection from roughly twenty-five paid social campaigns.
One offer. Three ways of saying it.
The interest-free credit offer is the single strongest lever a furniture retailer has, and it's usually wasted on a headline that just states the terms. Rather than build one execution, I built the campaign as a test: three message variants, each attacking a different reason someone hasn't bought yet, run across three room sets so the visual stays constant and the message is the variable. Nine cuts, all 9:16, all around ten seconds.
Every cut on this page is edited to a soundtrack. They play muted here on purpose: it's how the feed plays them, and none of them needs the audio to make sense.
Read across for the same message in three rooms. Read down for three messages in the same room. That's the point: the paid team can attribute a difference to the copy or to the set, not to both at once.
Twenty-five campaigns. Every one of them more than once.
Most of the work is promotional and repeats on a calendar. Some of it isn't: when a brief needs a room that doesn't exist, it gets built.
Both of these began as a single static product photograph and ended as a fully three-dimensional environment. The furniture is the real product; the room, the light and the camera move are all built. A brief that would otherwise need a location, a crew and a shoot day gets made at a desk, and the product still looks exactly like the product, because it is the product. These two took longer than anything else here and landed best with the team.
One concept, every placement.
Each promotional campaign ships in at least two ratios. The vertical is not a crop of the square: the type reflows, the product re-frames, the offer moves to where the thumb is. Reformatting is where most retail paid social quietly loses its legibility, so it gets designed rather than exported.
Same campaign, five trend cuts.
Home Event SS26 ran as a set rather than a single film, one cut per interiors trend, so the paid team could point each variant at a different audience instead of hoping one execution covered everyone.
Not everything should look like an ad.
Polished animation wins some placements and loses others. Alongside the campaign work I shoot and cut lo-fi vertical content in the stores: handheld, staff on camera, native to the feed it's going into. The Corby outlet walkaround was shot in 4K vertical and cut into a set of separate entry points rather than one long film, so the paid team could test which opening actually held.
Corby Outlet. Five cuts from one shoot, 17–54 seconds, 4K vertical. Three rounds of edit before release.
Creator-style, made in-house.
A separate strand shot in the Reading store: colleagues presenting to camera, shot on a phone, cut to look like the content around it rather than the advertising above it. Practical formats — how to choose a rug size, armchair comparisons — plus a football tie-in that let a furniture retailer join a conversation it had no obvious right to.
Product video or animation? Stop arguing, build both.
A recurring internal debate: should product creative be filmed or animated? Rather than settle it in a meeting, I built the same three products each way so the difference could be looked at instead of asserted. This one was an internal exercise, not a published campaign, but it's the habit I'd bring to any performance team: when two opinions are cheap to test, test them.
Internal test, unpublished. Included because the method matters more than the artefact.
This work was built to be measured. The numbers aren't mine to publish.
On metrics: performance data for this work sits in Furniture Village's ad accounts and isn't mine to publish. Rather than pad this page with numbers I can't stand behind, here is what's verifiable: the work ran, on the channels named, at the volume named.
What I can talk through: how the variants were structured so a result could be attributed to one variable, which hooks I'd expect to win in this category and why, and what I'd change in the next round. Happy to go through it properly in conversation.