Read Time: 3 Minutes

Something broke in July of 2025.

Facebook Ads campaigns that had been stable for months started failing overnight not a gradual decline in performance.

Ads that converted well would suddenly stop performing.

New campaigns would either explode or collapse within 48 hours.

No middle ground. No predictability.

I kept thinking: "This will stabilize soon. Facebook always does this."

Three months later, I'm still seeing the same chaos across multiple ad accounts. Not all accounts, but many.

And I finally understand why.

It's not a bug. It's Meta Andromeda.

The Algorithm Upgrade Nobody Warned You About

In July 2025, Facebook quietly rolled out their biggest algorithm update in years.

They called it Meta Andromeda.

The promise? Faster, smarter, more personal ad delivery.

The reality? Chaos for advertisers still playing by the old rules.

Here's what happened:

Facebook solved their "retrieval problem."

For years, they struggled to match the right ad with the right person from millions of potential ad options. So generic ads worked. Big hooks that appealed to huge audiences got results.

"Hey, if you love steamy romance, you'll love this..."

Broad. Safe. Effective.

Not anymore.

Andromeda changed the game.

Now Facebook can actually find your exact readers in the sea of their 3 billion daily users.

But there's a catch.

The New Rules

The algorithm wants creative diversity.

Not three versions of the same ad with different hooks.

Not the same image with a different background color.

Real diversity.

Here's what that means:

Different concepts - varying structures, different angles, multiple hooks

Different formats - short videos, long videos, static images, carousels

Different everything - primary text, headlines, descriptions, visuals

The old playbook? Make one great ad and scale it.

The new playbook? Make 10-20+ diverse ads and let Facebook figure out who sees what.

Why Most Authors Are Fighting This

I spent three weeks refusing to accept it.

I kept doing things "the old way", forcing things to work.

Zero improvement.

Because I was still focused on the 10% that doesn't matter.

Campaign structure. Targeting settings. Budget allocation.

Meanwhile, I was ignoring the 80%+ that actually drives results.

The ads themselves.

What's Working Now

For the last few weeks, I've been rebuilding my entire Facebook Ads strategy around creative diversity.

A lot of what I tested, failed. But I'm now gradually starting to see some glimmers of hope. 

Here's what I've been testing:

• Primary Text that's 3 sentences vs. 3 paragraphs
• Headlines that ask questions vs. headlines that make statements
• Images with characters vs. images with just book covers
• AI-generated images vs. elements of the book cover
• Images that purposely look like ads vs images that "blend in" a little more
• Text only images vs colorful and vibrant images with characters

I'm letting ads run for 2 weeks now instead of 1 week to give the algorithm more time to figure out the right ad for the right person.

Same broad targeting audience. 

But massively different creative.

The Uncomfortable Truth

You can't out-optimize Facebook's algorithm.

They've spent billions of dollars on this. They process trillions of data points daily.

Every time you restrict targeting, you're limiting what the algorithm can do.

Every time you create complex campaign structures, you're confusing it.

Stop it.

The algorithm knows your readers better than you do. Let it work.

Your only job? Give it diverse creative to work with and give it time.

What This Means for You

If you're spending hours in Ads Manager tweaking settings, you're doing it wrong.

If you're recycling the same three ads for months, you're leaving money on the table.

If you think one perfect ad will solve everything, you're going to be disappointed.

The authors winning with Facebook Ads right now are doing one thing differently.

They're treating creative testing the same way authors running BookBub Ads treat comp author testing.

Volume. Diversity. Data-driven decisions.

Launch 10-15 completely different ads.

Let them run for 2 weeks.

See what works.

Make more of what works.

Cut what doesn't.

Repeat.

My Challenge to You

For the next 30 days, don't touch your targeting settings.

Keep it simple: location, age, gender. That's it.

Instead, spend that time creating new ads:

• Different images
• Fresh Primary Text angles
• New Headline approaches

One month. Zero technical changes. Just creative.

I'd bet money you'll see better results than the last three months of optimization combined.

Because Andromeda doesn't care about your carefully crafted audience segments.

It cares about having enough diverse creative to match with the readers who actually want your books.

Give it what it wants.

Thanks for reading, and have a wonderful weekend.

See you next week.

To Your Success
– Matt