Hire an AI ad test planner and lessons reviewer

Everything it can hand back

Your task sets the actual outputs. The specialist picks the methods that cover them.

• A gap map and five untried ideas, each with a test question
• A test campaign layout with pack names, budget split and scale rules
• A ranked list of the test ads and what the winners share
• A lesson written as a rule to test next, with how strong the evidence is

How it does the job well

• Write for one named audience and one action. Each asset has a single job.
• Use the customer's own words from reviews, calls or surveys before inventing language.
• Every claim needs proof the business can show. Leave unproven claims out or mark them.
• Judge results against the business goal, such as profit or qualified leads, not clicks alone.

Mistakes it avoids

• Generic copy that could fit any brand.
• Comparing periods or campaigns that are not comparable, such as different dates, budgets or tracking.
• Calling a test winner before there is enough data.

What a good result looks like

An example, not your task: You have months of ads and no record of what they taught.

• outputs/step-01-untried-ad-idea-finder.md. Evidence grade, a gap map across five lenses, five ranked gaps, and five ideas each with a test question.
• outputs/step-02-meta-test-campaign-planner.md. Campaign structure, ad packs, budget split, CBO or ABO choice, and the keep, cut and scale rules.
• Pause. You launch the test, leave it alone for the read window, then export the results.
• outputs/step-03-winners-and-losers-ad-comparer.md. Ranked ads, a checklist per ad, what winners share, a rule to test next, and how strong the evidence is.
• HAND-BACK SUMMARY. Each step's file, any step not run and why, and every MISSING line in one list.

How the work is checked

• Each idea differs from the others on at least two lenses and says what you learn if it wins.
• The test layout fits the daily budget and never adds ads to a winning set.
• Step 3 used the results export from the pause, and says how many ads the lesson rests on. Under ten, it is called a hint.
• The rule to test next is one sentence with its evidence.
• States which checks passed, failed or could not run, and names missing inputs.

These checks have not been run against a model yet. Your AI uses them to check its own work.

A second opinion

Before it hands back, your AI reviews the draft the way John Caples and Eugene Schwartz would. This is a way of reviewing the work. It does not mean they wrote or endorse it.

Capabilities

• Find the ideas you have not tried. Records each past ad, looks for gaps across moment, value, buyer knowledge, opening and form, ranks the five biggest and writes one idea per gap.
• Lay out the test campaign. Reads the ideas, groups the new ads in packs, picks the budget type, splits budget by stage and writes the keep, cut and scale rules.
• Pause. Launch and wait. You run the test without changes for the read window, then export the results.
• Learn from the winners and losers. Starts from the results you export after the run, and reads them against the ideas and the pack names and rules. Puts a number on each ad, ranks them, fills a checklist for each, finds what winners share, checks it is not a place effect and writes the rule.

How the role works

1. Checks your task, context and available tools.
2. Picks the methods needed and prepares a work plan.
3. Carries out supported tasks, with each method's limits and review gates.
4. Checks the real deliverables and corrects supported faults.
5. Hands back files, evidence and unresolved gaps.

Your specialist's working setup

Download the base specialist without a task form or account. README.md and SETUP.md explain how to use the folder. SOUL.md defines its scope. INTAKE.md asks missing questions inside your AI. PLAYBOOK.md, SKILLS.md, BEST_PRACTICES.md, CHECKS.md and HANDOFF.md guide the work and its checked deliverables.

Complete associated skills, supporting files and licences are included. Choose only the methods needed for the actual task. Keep context, working files and accepted feedback in your AI project.

Use this specialist through MCP or the API

Call prepare_specialist with role_id plan-ad-tests-run-them-and-keep-what-they-teach only. No task is required to load it. Your AI asks what job you need done, what you want back and which relevant context it can use.

For task-led hiring, call find_specialist, explain the suggested specialists and let the user accept or change the team.

Complete ZIP: Download base specialist. API: GET /api/v1/specialists/plan-ad-tests-run-them-and-keep-what-they-teach/pack. The host AI supplies its own model, tools and execution; a pack does not connect accounts or run a task.

Inputs and helpful context

The specialist starts from the task brief. Only active methods require their inputs.

For the primary method, when used

• Your brand file
• Your past ads and their results (spend, sales or replies)
• Your budget, your target cost per sale and the number of products to test (for the layout)
• How many days you will wait before reading the test (the read window)
• After the run, the export of results for the test (you add this at the pause)

Tools and access

• Meta Ads Manager: An export by ad gives the 'measured' grade, with spend and results for each ad over the dates you pick.
• Meta Ad Library: Shows the ads your page runs now, with start dates, when you have no export to work from.
• Google Sheets: Keeps one row per past ad, so each lens and the moments grid are filters you can check.
• Meta Ads Manager: It holds the ad set spending limits, the exclusions and the campaign settings the plan is built around.
• Any good chat AI: Any current chat AI can build the plan from your numbers. Check each setting against the live menu.
• A spreadsheet: Holds the checklist as a table and counts values for each group.
• Your ad platform or email tool: Exports replies and spend by ad.

Functions and deliverables

• Ad test planner and lessons reviewer: Share product facts, past ads and results. Get test ideas, a campaign plan and a lessons review after real results arrive.
  • A gap map and five untried ideas, each with a test question
  • A test campaign layout with pack names, budget split and scale rules
  • A ranked list of the test ads and what the winners share
  • A lesson written as a rule to test next, with how strong the evidence is
• Untried ad idea finder: Find untried ad ideas for your brand. Paste your past ads and brand file. You get a gap map and five ad ideas, each one a clear test.
  • A record of every past ad, with each result graded measured, reported or seen running
  • Five lenses and a list of your proof, showing what's covered, thin or open
  • The five biggest gaps, ranked, plus the proof worth collecting
  • Five ad ideas, each with its test question, a five-link check, headlines and an opening
  • Any check the map still missed, one line each
• Meta test campaign planner: Share ad concepts, target cost per sale and budget. Get test groups, naming rules, settings checks and bounded spend plans.
  • A campaign and ad set layout sized to your store
  • A pack plan with ad names, minimum spend and removal date
  • A CBO or ABO call with the trade-off written out
  • A settings checklist, naming guide and rules for adding ads and scaling
• Ad results comparison analyst: Find what your best ads have in common. Give it results for five or more ads and get them ranked, with a rule to test and the counts behind it.
  • The ads ranked on a fair number
  • A completed checklist for each ad
  • What the winners share, with counts for winners and losers
  • A check that the pattern is not a place or season effect
  • A rule to test and a lesson record
• Brand profile builder: Share your website. Get a reusable profile covering your products, customers, voice and approved claims, with sources.
  • A folder of eight plain files, from a start file for your AI to a claims list and a change log
  • A tag on every fact, and every claim marked approved, pending or banned
  • A two-sentence diagnosis of your biggest gap, and a list of what's missing and who can answer it
  • A short slice for your next step, if you name one
• John Caples: Write headlines and plan a fair ad test through John Caples’s recorded approach. Lead with the reader’s gain and count replies by version.
  • Your problem worked on the way John Caples would: their questions, their method, their push-back
  • The answer or the work, in John Caples's voice, using their ideas by name
  • Anything that goes beyond their public record, marked
  • One suggested next step
• Eugene Schwartz: Match a headline or sales page to what buyers already want and know. Apply Eugene Schwartz’s recorded approach to research, claims and proof.
  • Your problem worked on the way Eugene Schwartz would: their questions, their method, their push-back
  • The answer or the work, in Eugene Schwartz's voice, using their ideas by name
  • Anything that goes beyond their public record, marked
  • One suggested next step

Tasks you can use it for

• You have months of ads and no record of what they taught.
• You are about to brief a new round of ads and want ideas that differ from past ones.
• A test ended and you cannot say why the winner won.

Limits to check before starting

• You have fewer than ten past ads with results: use the Untried ad idea finder on its own first, or plan one simple test by hand.
• You only want to read the results of a test that already ran: use the Ad results report writer.
• Your sales dropped and you need to know why: use the Ad sales drop analyst.

Best way to use this role

Give it one clear outcome, the actual files and a definition of done. Keep the work in one AI project so it can reuse checked context and earlier files.

• Download the handoff pack. It's a zip with HANDOFF.md and every skill file this workflow uses.
• Give the folder to Claude (in Claude Code, or as files in a Claude Project) and say "Follow HANDOFF.md".
• Export your past ads with dates, spend and results for the first step.
• At the pause, launch the pack and do not change it until the read window ends. Then export results and come back.

Read it as data

Read the complete role setup as JSON. Connect your AI to the AI Blocks MCP library and read aiblocks://roles/plan-ad-tests-run-them-and-keep-what-they-teach for its role instructions and pinned sources.

This is a prepared AI role. No task has run. Your AI carries out authorized work using its available tools and reports the actual deliverables and checks.