How I Automated 12 Copy Legends Into One Ad-Writing System

Powered by n8n and GPT. Writes, scores, and fixes ad copy on autopilot.

Ad copy makes or breaks your ads.

I once ran a $0 AI copy fix that 2x’d CTR and boosted ROAS by 48%, without changing the creative or targeting. Just better structure and emotional hooks.

So instead of asking GPT to “make this better,” I built a system that does it all for me.

It’s called the Ad Copy Engine.

I took 12 proven frameworks from 12 copywriting legends—Schwartz, Halbert, Ogilvy, Sugarman—and baked them into an n8n flow.

You fill in three fields:

• Product name

• Product description

• Target audience

The flow writes 12 ad variations, each in a different framework style.

Then each one gets scored from 1 to 10.

If it scores below a 9, it gets rewritten automatically.

Then all 12 final ad copies (scored, optimized, and labeled) are pushed to a Google Sheet, ready to deploy anytime.

The Ad Copy Engine, Visualized

What Each Part Does:

• On form submission
  You enter: Product name, description, and target audience.

• Copywriting Frameworks (Edit Fields)
  Loads 12 copy styles from legendary copywriters (e.g. Schwartz, Halbert, Ogilvy).

• Split Out → Loop Over Items
  Loops through each framework one by one.

• Generate Ad Copy
  GPT writes ad copy using that specific framework.

• Evaluate Ad Copy
  GPT scores the ad from 1 to 10 based on stopping power + buying intent.

• If Score < 9
  Routes low-scoring ads to the next step.

• Tweak Ad Copy
  GPT rewrites the ad (usually just the opening lines) to make it stronger.

• Google Sheets
  Logs all 12 ads—final copy, framework used, and score—into a clean spreadsheet

Under the Hood (On Form Submission)

To run the flow, you only need three things:

• Product name

• Product description

• Target audience

But here’s the key: your target audience has to be specific.

Don’t write “females, 18–65.”
Write: “First-time moms struggling with sleep after their newborn wakes up every 2–3 hours.”

The deeper the audience insight, the sharper the copy GPT will write. That’s how this workflow works its magic.

Under the Hood (Copywriting Frameworks Node)

This is where framework marketing kicks in.

Instead of asking GPT to "write better ads," I fed it 12 battle-tested frameworks each one mapped to a legendary copywriter.

The result? Structured output. Consistent angles. Scroll-stopping copy without guesswork.

Under the Hood (Generate Ad Copy Node)

Here’s what I think the anatomy of a good prompt should have:

• Role: Tell GPT who it is (here, a world-class direct response copywriter).

• Context: Tell it what it’s doing (writing ad copy using a specific framework).

• Constraints: Tell it how to behave (short, rhythmic, emotionally sharp, no markdown).

This node does all three. Every ad is structured, scroll-stopping, and emotionally tuned.

Under the Hood (Evaluate Ad Copy Node)

This is where GPT becomes your customer.

You tell it to think like your target audience—tired, distracted, mid-scroll—and evaluate your ad through their eyes.

Just a raw 1–10 score based on curiosity and scroll-stopping power.

Under the Hood (Tweak Ad Copy Node)

If the score’s below 9, this prompt kicks in.

It rewrites the copy like a seasoned copy chief. This fixes rhythm, tension, or emotional payoff without touching the core idea.

Output: Final Ads, Fully Scored and Stored

Each ad gets logged in a Google Sheet along with its framework, score, and whether it was tweaked.

Everything’s ready to deploy, A/B test, or swipe for future campaigns.

That’s all I have for this workflow!

I test AI workflows on real marketing problems.

And turn the useful ones into 10x Playbooks.

So business owners and solo marketers doing it all themselves can skip the guesswork… and copy what actually works.

→ I drop one of these 10xPlaybooks 🚀 every week, don’t miss the next.

John