From boosted posts to a tested ad plan
How a subscription razor brand replaced guesswork with a scored, ready-to-run ad strategy, built by a team of AI specialists in one working session.
01 At a glance
The short version
A direct-to-consumer razor subscription brand was spending on ads without a plan behind them. We set up an AI ad strategy team and gave them one.
Subscription razor brand
Sells blades and shaving products online, delivered monthly. Competes with big supermarket brands on price and convenience.
Spend without a strategy
Ads went out to broad audiences with no clear buyer, no funnel, and no test plan. Getting an agency strategy would take weeks.
Five AI specialists at once
Audience, creative, funnel, competitor, and budget agents analysed the brand together and produced one scored strategy.
Acquisition cost halved
Targeted buyer profiles, 25 tested hooks, and a structured funnel doubled monthly subscribers from 120 to 240 on the same ad budget.
The shift
Before: spend first, work out why later. After: knowing who to target, what to say, and where to put every dollar before scaling.
02 The pain
Ads were running.
A strategy wasn't.
The problem wasn't the ads themselves. It was everything that should have come before them.
Targeting everyone
Ads reached broad audiences instead of a defined buyer. Money went to people who were never going to subscribe.
Budget split by habit
Spend was spread across platforms without a reason. No one could say which platform or customer type deserved the most.
Generic copy
Ads described the product, not the buyer's frustration. Nothing stopped the scroll, and there was nothing to test against.
No funnel
Every ad asked for a sale. There was no plan for warming up new people or bringing back visitors who left.
Strategy was slow and costly
A proper agency strategy meant weeks of research and a large fee before a single ad could be tested.
03 The solution
An AI strategy team,
working in parallel
We set up a team of AI specialists inside Claude Code. One command, /ads strategy, sent the brand's website to five agents at once. Each worked from the same briefing, so their outputs lined up.
Read the brand's site, worked out the business type and industry, and picked the likely ad platforms.
Five agents analysed audience, creative, funnel, competitors, and budget. Each returned a score out of 100.
Scores were weighted into one Ad Readiness Score, a grade, and a client-ready PDF report.
Human review stayed in
The AI produced the first draft. Our team checked the buyer profiles against what the brand knew about its customers, trimmed the budget plan to a sensible test amount, and checked copy against each platform's ad rules before anything went live.
04 Implementation
What we ran,
and what it gave the team
The full strategy came first. Then focused commands went deeper on the parts that needed it.
| Command | What it gave the team |
|---|---|
/ads strategy | The full five-agent strategy, readiness score, and action plan |
/ads audience | Detailed buyer profiles with ages, interests, platforms, and targeting settings |
/ads competitors | What rival razor brands run, where they spend, and the gaps left open |
/ads hooks | 25 opening lines across five angles, with variations to test |
/ads copy | Headlines, ad text, and calls to action written for each platform |
/ads funnel | Four campaign stages, from first seeing the brand to coming back to buy |
/ads budget | A monthly split across platforms with clear cost targets and returns |
/ads landing | A review of the landing page: what blocks sign-ups and what to rewrite |
/ads testing | A test plan: what to test, for how long, and when to pick a winner |
/ads report-pdf | A six-page strategy report for the brand's team |
Installed the ad skills and agents in Claude Code and ran the full strategy.
Ran the focused commands and reviewed every output by hand.
Walked the brand's team through the report and agreed the first tests.
05 Under the hood
The script behind
the score dial
This Python script builds the finished PDF report. The part below draws the Ad Readiness Score dial on its first page.
Sets up a blank drawing and finds its centre point.
Stacks three circles to make the ring. Its colour follows the score: green at 80+, blue at 60+, amber at 40+, red below.
Writes the baseline score of 79, with "/ 100" underneath. The same dial rose to 87 after 90 days of improvements.
06 The strategy
A clear picture of
where the brand stood
| Category | Weight | What it checks |
|---|---|---|
| Audience clarity | 25% | How well the buyer is defined, and how precise the targeting is |
| Creative quality | 20% | Hook strength, copy, visual ideas, video scripts |
| Funnel architecture | 20% | Campaign stages, retargeting, path to purchase |
| Competitive position | 15% | What sets the brand apart, and gaps in the market |
| Budget efficiency | 20% | Platform mix and target cost per click and per sale |
"The professional groomer"
Male, suburban, 28 to 40, earning $75k to $150k. Found on YouTube, Instagram, LinkedIn, and Reddit. The plan puts 30 to 35% of new-customer spend here.
Four funnel stages
Awareness, consideration, conversion, and retargeting. Each has its platform, ad types, success measures, and share of budget.
A sharper message
The Meta ads led with price pain: "Stop overpaying, start shaving smarter." Every platform got its own headline, text, and call to action.
$10,000 a month, with reasons
Split across four platforms, with Meta taking 50%. Each line details target reach, clicks, cost per sale, and return.
07 Creative
Opening lines built
around real frustrations
Generic copy was one of the biggest gaps. The hooks command gave the team 25 opening lines across five angles to test.
Pain / problem
Name a frustration the buyer already feels.
Curiosity
Open a gap the viewer wants closed.
Social proof
Lean on numbers, reviews, or authority.
Contrarian
Challenge what the category assumes.
Urgency
Give a reason to act now.
Testing plan
Test angles first, then variations, then scale the winners.
The standout: "The locked case"
Most razor buyers know the moment: waiting for a store worker to unlock the shelf just to buy cartridges. The hook names it, then shows the alternative. It came with:
- Variations to A/B test against each other.
- Why it works: the frustration it speaks to, and why buyers recognise it straight away.
- Platform notes: for Meta, a split screen of the frustrated shopper beside a box on the doorstep, used as the first three seconds of video. Versions for TikTok, Google, YouTube pre-roll, and email subject lines.
08 The outcome
What changed,
and how it performed
The strategy turned each pain point into clear, measurable improvements across the campaigns.
| Area | Before | After 90 days |
|---|---|---|
| Targeting | Broad audiences, no defined buyer | 35% of new-customer spend focused on the highest-return buyer |
| Budget | Spread by habit | A $10,000 monthly split with a reason for every dollar, Meta leading at 50% |
| Creative | A few generic ads | 25 hooks tested, with the top 3 scaled across platforms |
| Funnel | Every ad asks for a sale | Four stages, including active retargeting for drop-offs |
| Strategy time | Weeks of agency research | A full strategy delivered in one working session, refreshed quarterly |
| Measurement | No way to track readiness | Readiness score lifted from 79 to 87 on the dial |
How the 90 days ran
Ran the five hook angles on small budgets. "The locked case" and price-pain hooks won.
Tested variations of winning hooks. Fixed landing page drop-off points.
Shifted budget to winning creative and the priority buyer, lifting the score to 87.
The key difference
Instead of burning ad spend on guesswork, every creative test had a clear hypothesis. Within 90 days, subscriber sign-ups doubled without adding a single dollar to the ad budget.
09 The numbers
The results
at $10,000 a month
Same budget, better ads. Over 90 days, tested hooks and sharp targeting doubled new subscribers and halved customer acquisition cost.
Halved by month 3, as clicks got cheaper and more visitors signed up.
Twice as many sign-ups by month 3, with no extra ad spend.
Based on lifetime value. Every $1 spent brought back $3.60.
Month by month
| Measure | Before | Month 1 | Month 2 | Month 3 |
|---|---|---|---|---|
| Ad spend | $10,000 | $10,000 | $10,000 | $10,000 |
| Click-through rate | 0.8% | 0.9% | 1.05% | 1.2% |
| Website visits | 6,667 | 7,500 | 8,750 | 10,000 |
| Sign-up rate | 1.8% | 1.9% | 2.1% | 2.4% |
| New subscribers | 120 | 143 | 184 | 240 |
| Cost per subscriber | $83 | $70 | $54 | $42 |
| Lifetime value earned | $18,000 | $21,450 | $27,600 | $36,000 |
extra subscribers over 90 days (567 against 360 at the old rate) for the same $30,000 spend. That was $31,000 in added customer lifetime value.