WeDiscover - homeTalk to us
Services - Creative

Creative that performs.

Overview

Creative at WeDiscover

Real customer insight in every idea, and every ad made to perform.

Work

From idea to feed

Crew Clothing

A colour palette for every season

Summer is often remembered through colour. We showed Crew’s summer range as curated colour palettes, so each ad gives styling ideas and shows how pieces work together. We tested women’s, men’s and mixed versions of the carousel, and the idea can return each season with a new palette.

Aspria

Every benefit in one carousel ad

Aspria’s clubs offer spa, fitness and flexible work spaces, but a single-image ad could show only one of them. We gave each benefit its own carousel card, in Aspria’s premium style. Each ad opens in the local language, so it speaks to people who live near the club.

How we do it

Creative technology

The tools and methods behind the work: how we make creative at scale, see what works in it, and prove that new creative adds sales.

01

Production at scale

One idea, made for every placement, size and audience, without starting again each time.

Design systems
We build each brand’s ads from components in Figma. A new size, format or offer is a change to the system, not a new design, so the work stays on brand as it scales.
Dynamic creative
For catalogues and prices that change, ads are made from a product or price feed (DCO), so the copy and the offer in each ad stay current.
Versions in batches
Sizes, languages and offer variants are made in batches from the master design, and a designer checks each batch before it goes live.
02

Creative analytics

We read results by what is in each ad, not only by which ad won.

Ads tagged by feature
Every ad is tagged by its hook, format, message and offer. The results then show which features work across many ads, and that shapes the next round of ideas.
The same data as your media
Creative results sit in the same data as your media and in the shared creative tracker, so the creative and media teams decide from the same numbers.
Fatigue, spotted early
We follow each ad’s performance over time and replace it before it wears out, not after the results have dropped.
03

Testing

We test where it matters, and claim an uplift only when we can measure it.

New ad sets
When we can split the audience, we run a split test: the old ad set against the new one. When we cannot, we use a before-and-after model with control groups.
New ads
We test five or more ads first and keep the best three or four, then measure the uplift with a causal model. We measure the whole ad set: Meta decides who sees each ad, so we never judge one ad alone.
Confidence agreed up front
We agree a confidence level with you before launch and claim an uplift only at that level. If the uplift cannot be measured, or is worth less than the test would cost, we monitor the ads and tell you.
FAQs

Questions about creative

Have a question that is not here? Talk to us.

Partner with us

Tell us what you want your marketing to do next.

Join our team

Join specialists in media, data, engineering and creative, and learn from all of them.