How I Built a B2B Ad Machine That Generates 7 Cold Ads (Copy + Visuals) in 10 Minutes

7 Cold Ads. 1 Workflow. Copy + Visuals.

When I joined Rezerv as an AI marketing strategist, one of my first tasks was to help launch paid acquisition from scratch.

New market, new offer, no historical data.

The playbook was clear: start with cold Meta ads.

But meta ads testing had changed. The emphasis is no longer on audience targeting. It’s on angles and creatives.

Which makes ad production the new bottleneck.

And that’s where things were too slow.

• Writing 5–7 variations for each concept.

• Briefing visuals. Waiting on revisions.

• Brainstorming suitable headlines.

So I built a fix.

A system that takes one ad concept and turns it into 7 cold-traffic ads, each with a different emotional hook, matching visual, Meta-ready copy, and headline. All generated in under 10 minutes.

The B2B Cold Traffic Ad Generator (Visuals + Copy)

What Each Part Does:

1. On Form Submission
   You enter: Brand name, ICP, key pain point, and a sample visual (e.g. website screenshot or social post).

2. Extract Brand Identity

   GPT analyzes the visual and extracts: HEX colors, fonts, and visual tone.
   Outputs a mini style guide in JSON format.

3. Load B2B Ad Angles

   Pulls 7 proven cold traffic frameworks (Pain, Enemy, Curiosity, Cost Framing, Before/After, Switch Trigger, Identity).

4. Split Out → Loop Over Items

   Loops through each angle one by one.

5. Generate Ad Copy

   GPT writes:

   • Visual overlay text for generated image (Audience callout + hook + CTA)

   • Meta ad copy + headline

6. Generate Visual

   Generates an on-brand ad image that matches the emotional angle and overlay text.

7. Upload to Google Drive

   Each image is converted to file and uploaded for access.

8. Google Sheets
   Logs all 7 ad variants each with its copy, Meta headline, emotional angle, and image URL into a clean spreadsheet for launch and tracking.

The B2B Ads This Workflow Generates

Generated Ad Copies + Headlines

The ad copies and headlines are crafted using voice of customer (VoC) inputs which are real phrases your target audience actually uses when describing their pain.

This makes each ad more emotionally compelling because it sounds like them, not you.

Generated Ad Copies + Headlines

Generated Images (Using gpt-image-1)

The visuals are generated by first referencing a specific ad style. In this case, I used bold billboard-inspired layouts that I think will work well for cold traffic.

Then, the system applies the brand’s extracted JSON visual identity (colors, tone, font cues) to ensure each image feels custom and on-brand.

Generated Images

How I Actually Use This System to Test ICP Pain Points

Creative testing isn’t random. It should start with your ICP’s top pain points.

For Rezerv, we identified 3 core pains:

1. Poor customer support from legacy competitors

2. Existing software is bloated and overpriced

3. No support for local payment systems

Each pain gets fed into the B2B Ad Generator → which spins out 7 variations per pain, using frameworks like Pain, Enemy, Curiosity, Cost Framing, Before/After, Switch Trigger, Identity.

That’s 21 cold ads, each one tied to a real-world frustration.

Then we run them.

Not to find a “winning ad.”
But to find the highest-performing pain point (measured by Outbound CTR and spend efficiency).

Once we know which pain hits hardest, that’s the variation we optimize with better visuals, stronger proof, refined copy.

This way, every ad we improve is built on validated pain, not gut feel.

The Cost 💰 & Time ⏰ Economics

Let’s talk cost:

• $0.20 USD per image generation

• Works out to $1.40 USD per run

• 3 runs (for 3 pain points) = $4.20 USD

Here’s what creating just one ad concept variation normally looks like:

• 15–30 mins researching writing copy (angle, hook, overlay text, caption, CTA)

• 5–15 mins tweaking with a designer (visual direction, brand colors, overlays)

• 5–10 mins going back and forth on revisions

• 1-2 mins logging everything into Google Sheets: copy, headline, image links, angles

That’s easily 30-45 minutes per ad concept variation, assuming nothing gets blocked.

With this system, you get 7 ad concept variations in under 10 minutes per run.

Wrap Up: AI Marketing = Bottleneck Removal

I built this workflow to solve a real bottleneck I was facing: turning insights into cold ads fast.

If you ever need this system, drop me a connect on LinkedIn, I will be happy to send it over.

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