Six months ago, we were spending hours crafting ad copy for our clients.

Testing different headlines, trying various angles, rewriting descriptions over and over. Our creative process was slow, and honestly, hit-or-miss.

Then ChatGPT changed everything.

Not in the way most agencies think – by replacing human creativity. But by amplifying what we already know about our clients’ customers.

The Breakthrough That Changed Our Agency

It started with a struggling e-commerce client. We’d been running Facebook ads for their skincare brand for three months with mediocre results. 2.1x ROAS, decent but nothing special.

Our usual approach: brainstorm headlines, write a few variations, test them for a week, pick the winner.

This time, we tried something different. We fed ChatGPT everything we knew about their customers:

• Customer reviews and testimonials
• Support ticket complaints
• Survey responses about skin concerns
• Competitor research

Then we asked it to write 50 different ad angles.

The result? One of those AI-generated angles hit 6.8x ROAS in the first week.

What We Learned About AI + Human Creativity

ChatGPT doesn’t replace our strategy. It amplifies it.

We’re still the ones who understand customer psychology, platform algorithms, and what drives conversions. But AI helps us explore way more creative directions than we ever could manually.

Here’s our process now:

Step 1: Feed AI Real Customer Data We don’t just ask for “skincare ad copy.” We give it actual customer language, pain points, and desires from reviews, surveys, and support conversations.

Step 2: Generate Volume Instead of writing 3-5 ad variations, we generate 30-50 options across different emotional triggers, benefits, and angles.

Step 3: Human Curation We pick the most promising concepts based on our experience with what works on each platform.

Step 4: Test and Scale The winners get budget, the losers get killed fast.

The Results Speak for Themselves

Since implementing this AI-enhanced process:

• Client A (Home Services): Conversion rate jumped from 3.2% to 9.7%
• Client B (B2B Software): Cost per lead dropped from $340 to $127
• Client C (E-commerce): ROAS improved from 2.8x to 7.2x

The difference isn’t just in performance – it’s in speed. We can now test more angles in one month than we used to test in six months.

What Most Agencies Get Wrong About AI

They either ignore it completely or try to replace their entire creative process with it.

Both approaches fail.

Ignoring AI means you’re stuck with the same creative limitations you’ve always had. You’re competing against agencies that can test 10x more concepts than you.

Replacing humans with AI gives you generic, templated copy that sounds like every other AI-generated ad on the platform.

The winning approach combines AI’s ability to generate volume with human expertise in strategy and curation.

Our Specific AI Workflows

For Facebook Ads: We prompt ChatGPT with successful ad frameworks, customer language, and specific pain points. Then we ask for variations that tap into different psychological triggers – fear, desire, social proof, urgency.

For Google Ads: We feed it high-converting search terms and ask for ad copy that matches specific search intent. The AI helps us create dozens of responsive search ad combinations we’d never think of manually.

For Creative Testing: We use AI to brainstorm entirely new angles based on customer data, then test these against our traditional human-written ads. The AI angles win about 60% of the time.

Why This Matters More Than Ever in 2025

Ad platforms are getting more competitive every month. The agencies that can test more concepts, faster, will dominate.

But speed without strategy is just expensive noise. That’s why the combination of AI generation and human strategy is so powerful.

We’re not just writing more ads – we’re writing better ads based on deeper customer insights, then testing them at scale.

The Future of Ad Creative

This is just the beginning. We’re already experimenting with AI for video scripts, image concepts, and landing page copy.

But the core principle stays the same: AI amplifies human expertise, it doesn’t replace it.