Process

Start with audiences.
Scale with ads.

Meta now demands ads targeted at much more focused audiences and niches: people looking for a specific product that serves a specific purpose. Our planning process is how we make more of the right types of ads, at scale.

Stage 01 · Planning

Granular audience personas, two positions each

Who are your customers?
What do they want to achieve?
How can your product help them do this?

Planning starts by defining granular audience personas rather than interest categories.

For each persona we identify two positions: the problem they're looking to solve and the outcome they're looking to achieve, and then the product purpose: how your product addresses both.

That gives us, for a typical brand, five audience personas, each with a problem to solve, an outcome to achieve and a product purpose that connects them.

Every ad we make from here has one of those as its brief.

Persona

Who they are and the relationship they have with your product today.

Position 1 · Problem

What's going wrong for them that they want solved.

Position 2 · Outcome

What they're trying to achieve, in their own words.

Product purpose

How your product addresses the problem and delivers the outcome.

Persona (CabinZero example)	Position 1 · Problem to solve	Position 2 · Outcome to achieve	Product purpose
Fee-averse budget flyer	Gets stung by airline bag charges at the gate	Fly for the fare they paid, nothing more	Sized to every major airline's cabin limits
The over-packer	Never fits a week into hand luggage	Five days, one bag, no hold luggage	Capacity and compression without breaking size limits
City-break regular	Wheelie cases on cobbles and stairs	Walk off the plane and straight into the city	Lightweight backpack that carries like luggage
Parent buying for a teenager	Kid's bag gets lost, broken or refused	A bag that survives the trip and the airline	Durable, trackable, with a warranty
Long-haul, multi-market	Bags fail after a couple of trips	One bag for years of travel	Built to last, localised for JP / AU / MY

Output of planning: 5 personas × 2 positions, each with a product purpose. This is the brief for every ad in the sprint.

Stage 02 · Ad types

10 ad types that we know work, because we made them

We take each persona's two positions and product purpose and apply our ten ad types to them. That is where the volume comes from: not more versions of one idea, but ten structurally different ads per position. It's how we get to 20, 40, 60 ads in a very short space of time.

5audience personas

2positions each

10ad types

100distinct ads from one plan

Show don't tell

Descriptions make the viewer imagine, and imagining is work. A demonstration removes the work entirely: they see the outcome and believe it in about two seconds. It works with the sound off, and it runs in every market you sell in.

Output benefit

Not the product, not the ingredient: the end result your customer actually gets. Most small brands instinctively advertise their features. Output Benefit ads lead with the life the product produces, in the first three seconds.

Solution problem

Opening on the problem asks a scrolling stranger to sit in something uncomfortable. Lead with the solved state and you earn the three seconds you need to explain what it was solving.

Peer to peer

The same sentence is worth more from someone with nothing to gain. It sits in the feed as content rather than advertising, and different faces fronting the same script is, to both the viewer and to Meta, a genuinely different ad.

Founder story

They can clone your demo, your offer, your comparison table. They cannot clone why you started. The founder is free talent and the story is a free script, and audiences reward the honesty a big brand can't fake.

Objection

The objection is already in their head. Saying it out loud first moves it from their side of the argument to yours, and the proof that follows lands as evidence rather than defence.

Us vs them

Punch at the category default, not a named rival. You're not asking them to switch supplier; you're asking them to stop paying for something that stopped making sense.

Offer

A discount trains people to wait for the next one. Added value raises what the thing is worth instead of lowering what it costs, and it lifts AOV rather than cutting into it.

Challenge value

Price is only a problem in the absence of a comparison. Say the number out loud, then stack the justification (the build, the guarantee, the thing the cheap version can't do) until the price reads as evidence of quality.

Social proof

At the end of the consideration cycle people aren't looking for a new argument, they're looking for permission. Where Peer to Peer is one person recommending, Social Proof is the weight of the crowd.

10 ad types that we know work, because we made them.

Stage 03 · Scaling production

A production team built for volume

Image ads, video ads, UGC and our own video creation, produced by a team of more than 25 video editors based around the world who can make ads at scale. The core team plans and briefs; the flex team scales up for the sprint and back down between them, so you're only ever paying for output.

25+video editors, worldwide

12videographers

5designers

2copywriters

Case in point · Big Potato Games

A six-week sprint. 100+ unique video ads. $3.5m generated.

Week 0The brief, 1 November: find winning ads that can scale an $800k budget before ChristmasThe brief

Weeks 1–2Personas and positions for Herd Mentality; hooks, angles and scripts, then the first production wave50+ ads

Weeks 3–4Second wave briefed off the first fortnight's data, across all ten ad types100+ ads

Weeks 5–6Scale the winners into Thanksgiving and BFCM$3.5m generated

#1 on Amazon in UK + US by Christmas · £500k in a single day · CPA −45% vs 2024

From the sprint · Herd Mentality9:16

Big Potato

Want to see this run on your brand?

Send us the ad account and the asset library. We'll map the personas, size the sprint and tell you how many ads it needs.

Get more ads →

Start with audiences.Scale with ads.