£500. 12 AI-drafted headlines, scored, with operator commentary. Ship in 5 working days.

AI Headline Lab

AI Headline Lab is a £500 one-off engagement: 12 AI-drafted headline alternatives for your homepage hero, scored by predicted CTR and conversion, with operator commentary. Delivered in 5 working days.

£500. 12 AI-drafted headline alternatives. Operator-graded. Ship in 5 working days.

Your homepage hero headline was written once, by your designer or co-founder or marketing manager, two years ago. It has not been tested since. The visitors arriving today read a sentence written for the visitors arriving in 2024.

AI Headline Lab is the fastest, cheapest way to find out which headline you should ship instead.

What you get

A single PDF deliverable containing:

12 AI-drafted headline alternatives for your homepage hero (or any single landing page you nominate)
Each headline scored on predicted click-through rate (estimated against your current baseline) and predicted conversion rate (estimated against published category benchmarks)
Operator commentary on each headline explaining the underlying audience hypothesis, the angle being tested, and the recommended A/B test setup
A single recommended ship, the headline I would deploy first if you only ran one test
A 15-minute call to walk through the deliverable, answer questions, and confirm the ship

No retainer. No follow-up sequence. No upsell. £500 flat. Done in 5 working days.

Who this is for

This engagement is built for one specific buyer profile: an operator who already knows the headline matters, doesn't have time for a full retainer, and wants a fast evidence-grounded answer to one specific question.

The buyer typically:

Runs an ecommerce or SaaS site doing £50K-£500K monthly revenue
Has not tested the homepage hero in 12+ months
Is paying for paid traffic to a hero that has never been audited against intent
Wants to ship a winner this fortnight, not enter a 90-day testing cycle

If that's not you, the Sprint, Growth, or Scale tier is the right starting point. AI Headline Lab is deliberately narrow.

How it works

Day 1. Brief. You send me three things via a short Loom or written brief: (1) the URL of the page, (2) your audience hypothesis (who's buying and why), (3) any existing headlines you've already tried. I do not need a full creative brief. The audience hypothesis is what matters.

Days 2-3. Generation. I generate 12 AI-drafted alternatives across four angle clusters: outcome-led, identity-led, fear-led, curiosity-led. Each cluster gets three drafts. The drafts are operator-shaped, not raw AI output. I rewrite anything the AI produces that fails the voice and brand-fit filter.

Day 4. Scoring + commentary. Each alternative is scored against two predicted metrics: click-through rate and conversion rate, each estimated against your baseline and published category benchmarks (Build Grow Scale, Baymard). Operator commentary annotates each headline with the audience hypothesis, the angle, and the recommended test setup.

Day 5. Delivery + call. PDF lands in your inbox. We hop on a 15-minute call to walk through it, answer questions, and confirm the recommended ship. After the call, the engagement is complete.

Why operator-graded matters

A purely AI-generated set of 12 headlines is what every DIY tool produces. The 4-7% lift band documented in Build Grow Scale's 2026 research across 347 stores is what those tools deliver. AI Headline Lab adds the operator layer, 13 years of CRO pattern recognition picking which alternatives are worth scoring and which are obvious losers.

The headlines that make the final 12 are the ones that survive the operator filter. The headlines that don't make it are documented in a discarded-with-reason appendix so you can see why a tempting AI draft did not ship, usually because it failed an audience-intent check or violated a voice-fit constraint.

This is the same operator-led methodology that produces the 28-34% expert-guided lifts in The 4-to-34 Gap, compressed into a 5-day £500 deliverable.

Real-engagement reference

The headline-testing layer is one of the disciplines inside OperatorAI. On full retainer engagements, headline tests run on a quarterly cadence and ship 30+ A/B experiments per quarter with 99% statistical significance gating (The 99 Rule).

A specific reference: on the Enzymedica Black Friday 2021 engagement, three compounded CRO wins moved baseline conversion 3.4% to 16.9% (5x lift) on the highest-stakes day of the year. Headline testing was one of the three compounding wins. AI Headline Lab packages the headline-testing layer as a standalone, low-commitment entry point.

What this is not

Not a full CRO audit (that's the Sprint tier at £2,500)
Not ongoing optimisation (that's the Growth tier at £2,500/month)
Not a multi-page test programme (that's the Scale tier at £5,000/month)
Not a copywriting service for the rest of the page
Not a brand voice review
Not a free audit (the free audit is at /audit)

The boundary is deliberate. £500 buys 12 headline alternatives, scored, with commentary, and a recommended ship. That's the whole engagement.

Pricing

£500 + VAT (where applicable). Fixed price. Paid via bank transfer or card on engagement start.

No retainer lock-in. No follow-up sequence unless you ask for one. No first-month discount that ratchets up. The £500 is the £500.

How to book

Email chris@gogochimp.com with the subject line AI Headline Lab and the URL of the page you want tested. Or use the booking form on this page. I'll come back the same working day with a brief template and an engagement start date.

If you've already had a free audit at /audit, reference the audit and I'll skip the brief step.

FAQ

How is this different from using ChatGPT to generate 12 headlines?

You can absolutely generate 12 headlines yourself with any LLM. The difference is the operator layer: which 12 to keep, how each is scored against your audience hypothesis, and what to ship first. The 4-7% lift DIY tools produce reflects the absence of that layer. The 28-34% expert-guided lift The 347 Method documents reflects its presence.

What if I want to test more than the homepage hero?

The £500 covers one page. Each additional page is +£300 if added at engagement start. For multi-page testing, the Sprint tier at £2,500 covers up to 5 pages.

Do you guarantee a lift?

No. The £500 buys you a tested set of 12 alternatives plus a recommended ship. The lift you see depends on your traffic, baseline, and audience match. Build Grow Scale's research suggests expert-guided AI testing produces 28-34% lift on average; that's the band to expect, but no specific result is guaranteed for a single engagement.