Most creative strategists have become so reliant on AI that the fundamentals - copywriting, real market research, the psychology of why someone actually takes action, have quietly stopped being the priority. From the profile to the proposal, minimal thought goes into WHY before the process gets automated.

Tools are only as good as the operator. I learned the discipline first and built the automation second, and that order is the entire difference.

Here's exactly what that looks like.

01 RESEARCH: read the audience, not the dashboard

Before a single ad gets written, I go and find out what your market actually says.

Review mining across your products and your competitors' - the wording people use when they're happy and when they're not

Forums, groups, comment sections and communities where the problem gets described in the buyer's own language

Your Meta data read properly: what's already working, what's fatiguing, and why

AI used where it earns its place - Gemini deep research to widen the net, never to write the thinking for me

You get ideal customer profiles built from evidence, a ranked list of the burning desires and objections that actually move people, and a vocabulary bank of the phrases your market already uses.

02 - MESSAGE: angles, not adjectives

Research is worthless if the copy doesn't land it.

An angle bank built directly off the research, each angle mapped to a specific avatar and a specific objection

Direct response structure - hook, mechanism, proof, offer, next step

Fundamentals studied from Halbert, Sugarman and Hormozi: writing to one person, keeping momentum line to line, and building an offer that's easy to say yes to

Copy written in the market's language so the reader feels seen rather than sold to

This is the part most people skip. It's the part that decides whether anything else works.

03 - DISTRIBUTION: structure built for how Meta works now

Andromeda changed how ads get distributed. Most accounts are still structured for the old system.

Account structure that suits Meta's retrieval system rather than fighting it

One testing ecosystem instead of phase testing with custom audiences

Creative targeting over manual targeting - the creative does the targeting

Deliberate creative diversity so the retrieval system has genuinely different work to push into different pockets of audience

Attribution discipline: new-customer CPA read across different attribution windows and verified in Triple Whale, rather than taking Meta's reported numbers at face value

Knowing exactly where Meta over-reports is a large part of what lets an account scale on real margin instead of a dashboard.

04 — PRODUCTION: volume without losing the thinking

I've built a set of Claude agents that carry my process — research in, finished statics out.

200+ statics per week when an account needs that pace

Formats produced deliberately: realism, billboard, claymation, retro editorial and native — each built for a different avatar, not chosen for variety's sake

Speed and cost of production treated as strategic advantages, not afterthoughts

"Statics only work at bottom of funnel" is a myth and a skill issue.

WHAT YOU GET

Market research dossier — review mining, forums, competitor teardown

Ideal customer profiles and avatar definitions

Angle bank with hooks mapped to avatar and objection

Static ad concepts and copy, production-ready

Account structure recommendations, or the rebuild itself

A testing roadmap with a clear order of operations

Reporting read against real numbers, not vanity ones

HOW IT RUNS

Week 1 — Research and ICP. Review mining, forums, Meta data. Avatar profiles and the language the market actually uses. Week 2 — Angles and first wave. Angle bank, scripts, and the first creative wave live, formats matched to each avatar. Weeks 3–4 — Structure and scale. Testing ecosystem rebuilt around what's winning, then spend goes behind the proven angle.

PROOF

Magic Tree Superfoods — took over an account turning £4,617.75/mo in conversion value at 1.90x. It now returns £105,983.30 across a rolling 30 days at 5.09x on £20,819.49 spend. A 23x lift.

One static, the "YOUR CHILD" angle — £9,957.28 returned on £1,804.43 spend. 5.52x, £11.72 per purchase, ~154 purchases from a single ad.

Jolie Skincare — scaled past $50,000/month. 2.75x account ROAS, 3.92x on retargeting.

Clean Green — the automation I built and own puts their statics at the top of the Meta Ads Library, which ranks by engagement.

Every contract I've closed has finished at 5.0.

Every figure above is backed by an untouched screenshot with the date range visible. Ask and I'll send them.

WHO THIS IS FOR

Ecom and DTC brands already spending on Meta who've hit a ceiling

Accounts where every ad looks the same and performance has flattened

Founders who want someone thinking ahead of the brief rather than waiting for it

WHO IT ISN'T FOR

Anyone looking for an order-taker

Anyone who wants fifty ads a week with no research underneath them

Anyone who wants the cheapest option rather than the one that works

In a field where the technology changes every quarter, I'll test anything and I'll happily be proven wrong by someone who knows more. That's an opportunity to improve while getting paid.

Send me a message and let's talk.