Most accounts I've been brought onto lately have the same complaint: more ads, same performance.

Every ad looks the same, targets the same buyer, and is the 13th reskin of a best performer from last year.

This usually happens because briefs are disconnected, multiple people are working different processes, or designers are simply left to fend for themselves, trying to guess what to build next.

Andromeda cuts tens of millions of live ads down to a shortlist before the auction starts. Near-duplicates don't get separate shots at that shortlist; they get collapsed into one entity and made to compete with each other.

Here's how to stop feeding it the same ad thirteen times.

One doc. Every decision made before a designer opens Figma.

For every variation: who it's for, what stage they're at, what's stopping them buying, what angle you're testing.

Do it right and every ad earns its slot. Skip it and half your "variations" are the same ad in a different shirt.

Build this out with your buyer personas; the more information for the delivery team, the better.

Here’s what’s included:

Basics: name, age, location, income bracket. Specific enough to be a person.

Pain points and desires: the problem in their words, next to what they actually want. Usually not the same thing.

Objections: what stops them buying even when they're interested.

Internal vs. external narrative: what they tell themselves privately, versus what they'd say to a mate down the pub. The gap is where your hooks live.

Scroll behaviour and common phrases: fast scroller or slow reader. Then the actual words they'd use, worth stealing for headlines.

Awareness journey: same person, five stages, from not knowing they have a problem to already using a competitor. Same persona, different ad, depending where they sit.

This information links straight into the planner, so every concept ties back to a real persona, not a guess.

My advice: keep it concise, easy to understand, and updated regularly.

Thirty-plus angles, based on common frameworks, not gospel. Rename them to match whatever your team already calls them internally.

Not a checklist. A reference, so you pick an angle on purpose instead of defaulting to whatever the last ad used.

This promotes structural variation, not a re-skin, and that's what Andromeda rewards.

This is where it comes together.

The previous tabs build the dropdowns; you can pick from what's already defined.

Select: target persona, awareness stage, pain points and objections. Sometimes it's both, sometimes just one. The more effort in the Personas tab, the more fun you can have here.

Outline the creative direction: media type and angle. Attach an existing ad from the wild, or knock up a quick wireframe for visual direction.

Write the content: headline, primary copy, CTA. This can be guidance or explicit, depending on your internal process.

Results fields: built in for tracking, and notes for designers and strategists to leave for whoever picks this up next.

Side by side, it's obvious fast. If two columns look the same, you've found the problem before media spend does.

It's bare-bones on purpose.

Enough structure to stop the reskin problem, nothing you have to fight to use.

Take it and build it out 10x further if that's what your team needs; add fields, tabs and whatever else to make it yours.

Use this as a starting point, not a rulebook.

The template only covers static concepts, but you could easily introduce hooks/scripts to the Planner tab.

Andromeda didn't make creative more important. It made planning it more important.

Plan the difference first. Design comes after.

Still briefing by vibes? Get in touch.