Writing Hooks: 20 Hooks in 15 Minutes

The hook is the most important part of any ad and with Claude you can produce 20 variations in 15 minutes instead of guessing with two or three

The hook is the most important part of any ad. It is also the part most people spend the least time on, writing two or three versions and going with whichever sounds best.

Professional ad teams test 15 to 20 creative variations per campaign. With Claude you can produce 20 hook variations in 15 minutes, run them against each other, and know which angle works before spending serious budget.

Why You Need 20 Hooks

The hook is a hypothesis. You are guessing which specific problem, desire, or curiosity angle resonates with your target audience right now.

Different people in your target audience have different reasons to care about what you do. One person is frustrated with manual processes. Another just got burned by an expensive agency. A third is embarrassed by how their website looks compared to competitors. A fourth is excited about AI tools but does not know where to start.

Each of these is a different hook angle. Each could be a winner or all four could underperform and the winner could be something you did not expect.

Testing is the only way to know. More hooks means more chances to find the angle that works which means a lower CPL.

The Three Hook Types

Question hooks create a pattern interrupt and make the reader answer a yes or no in their head.

Statement hooks are bold, specific, and sometimes provocative. They make a claim that is specific enough to be interesting.

Outcome hooks lead with the result and not the process. They show the end state.

In your batch of 20 aim for roughly 7 question hooks, 7 statement hooks, and 6 outcome hooks.

The 20-Hook Batch Prompt

Using @ADS.md, @SALES.MD, and @MARKETING.MD — write 20 ad hooks for

my paid advertising campaign.

What I'm advertising: [SERVICE OR OFFER FROM ADS.MD]

Platform: [Meta or Google — determines length and format]

Target audience: [WHO THEY ARE FROM MARKETING.MD]
Their biggest frustrations: [LIST 3 FROM YOUR DISCOVERY CALLS OR RESEARCH]

Their biggest desires: [LIST 3 — WHAT THEY ACTUALLY WANT TO ACHIEVE]

Write exactly 20 hooks split as:

• 7 question hooks (end with a question mark)
• 7 statement hooks (bold claim or specific observation)
• 6 outcome hooks (start with the result — what their life or business

looks like after)

Rules:

• Every hook under 125 characters (Meta mobile truncation limit)
• No generic phrases like "grow your business," "level up," or "take

your X to the next level"

• No jargon like "innovative solutions," "cutting-edge," or "bespoke"
• Each hook must be different — different angle, different emotion,

different focus

• Hooks should sound like a person said them and not like an ad
• Number each hook and label its type as Q, S, or O

After the 20 hooks tell me which 5 you would test first and why.

Before and After: Generic vs Strong Hooks

The difference between a weak hook and a strong one is specificity. Here are examples for a Claude Code builder who builds websites and tools for local businesses.

Generic hooks that do not stop anyone: "Professional website design services for local businesses." "Grow your business with AI-powered web solutions." "Do you need a better website for your business?"

Strong question hooks: "Still sending clients to a PDF form to book appointments?" "How many people visit your website every month and never come back?" "What's the one thing stopping your website from booking while you sleep?"

Strong statement hooks: "I built a dentist a self-booking website in one afternoon. He hasn't touched it since." "Most local business websites lose 60% of visitors before they ever scroll down." "Your competitor booked 3 new clients from their website this week. Yours didn't."

Strong outcome hooks: "8 new bookings last Tuesday. All from a website that runs itself." "From calling to confirm every appointment to zero phone tags. Same business, better site." "The plumber who gets 5x more online bookings than his competitors built his site in a day."

The generic versions could have been written about any business in any industry. The strong versions are specific, concrete, and make the reader think about their own situation.

Reviewing Claude's Output

After running the prompt read every hook out loud and apply three filters.

Filter 1: Specificity test. Could this hook have been written for any business? If yes it is too generic. Rewrite it with a specific detail.

Filter 2: Emotion test. Does this hook create a feeling such as curiosity, recognition, ambition, or mild anxiety? If it creates no feeling it will not stop a scroll.

Filter 3: Audience test. Would the specific person you are targeting recognise themselves in this hook? If not, who is it actually for?

Remove or rewrite any hook that fails two or more filters. Keep only the ones that pass all three.

Building Your Hook Bank

As you test hooks and see results record the winners and losers in ADS.md:

## Winning Hooks

Hook: "Still sending clients to a PDF form to book appointments?"

Type: Question

Platform: Meta

Result: 3.8% CTR · £18 CPL · Week 1

Note: Strong with dental and physio audience

Hook: "I built a dentist a self-booking website in one afternoon."

Type: Statement

Platform: Meta

Result: 2.9% CTR · £24 CPL · Week 2

Note: Works for awareness but not conversion

## Losing Hooks (kept for reference)

Hook: "Professional websites for local businesses"

Type: Statement

Platform: Meta

Result: 0.6% CTR · £89 CPL · Week 1

Note: Too generic — no emotional anchor

After three months your hook bank contains the patterns of what works for your audience. Claude can then write new hooks based on what has already proven to work.

Your Action Right Now

Run the 20-hook batch prompt. You need ADS.md started even as a partial version and your offer from SALES.md. Run it, get 20 hooks, and filter them down to your top 10.

In the next lesson you will take the best hooks and build complete Meta ad copy around them.