Prepared illustration. Fictional product. No connected account.

stephen@iamsbrown.com
00 / A working illustrationFictional retailer / Prepared with AI assistance

100 copy drafts.Make the strongest count.

One fictional product, 100 prepared copy drafts and the path from shortlist to headline test to next decision. Try the controls below.

100Unique copy drafts
10Messaging angles
50Headline hooks
2Copy structures per hook

The fictional brief

Fieldnote Daypack 01

A bag for work essentials and light days out. The task is to develop options for a purchase campaign, then choose a focused first test.

  • 22 litre capacity
  • Padded sleeve for a laptop up to 15 inches
  • Front zip pocket and two side pockets
  • Slate or olive
  • Example product price: £68
Inside the demonstration

This prepared library was written with AI assistance and is shown here as an example. The controls browse, shortlist and plan; they do not call a live AI model, connect to Meta or publish anything. These are copy drafts and creative directions, not 100 finished artworks. They demonstrate production and planning, not sales performance.

Explore the copy library
01 / Produce at scaleVolume becomes useful through selection

Inspect the pool. Pick the contenders.

Ten angles, five hooks per angle and two copy structures per hook. Shortlist up to six candidates. The next step uses two headlines in a focused test, with the other elements held constant.

Download all 100 as CSV

Planning CSV for this fictional example. It is not a platform upload file.

02 / Turn volume into a questionPlan a controlled comparison

Give the next test a clear job.

A large pool gives options. A useful test needs one question, checked claims and a budget and evidence threshold set before launch.

No candidates shortlisted.

Preview a headline test

This preview takes only the chosen headlines from the drafts. The body copy, visual direction, product and CTA below stay the same. It does not compare the full library drafts.

Before a real launch

Choose the business metric, audience allocation, spend cap, duration and decision rule using the account’s baseline and available traffic. Use a suitable controlled experiment when asking which headline caused a difference. Ordinary delivery across many ads is not an isolated A/B test.

This illustration specifies a method, not current Meta tool limits or a guaranteed result. Assets, account setup and measurement still need checking.

03 / Operate and improveA report leads to the next action

The report decides the next move.

In a live account, delivery, pacing and orders are monitored and the next change goes into the campaigns. This example has no account data.

Spend and pacingNot connected
Orders and acquisition costNot connected
Test resultNot run
Campaign actionExample only

Inspect a hypothetical reporting situation

These are decision examples, not observed account findings or simulated performance claims.

Book a discovery call

Bring your next promotion or launch. Channels, budget and success measures are set before anything goes live.