Why Your Remarketing Belongs in a Separate Campaign (Still in 2026)

I run remarketing in its own campaign.

Not as an ad set inside my prospecting campaign.

Not folded into an Advantage+ campaign with everything else.

Its own campaign, own budget, own creative.

I do this even for accounts spending only 1000 EUR a month.

I hear the counterargument often.

"My budget is only 1000 EUR a month. I don't need to complicate the structure."

You don’t.

You need 2 campaigns: Prospecting (cold) and Remarketing (warm).

That’s not complicated, isn’t?

Here's what happens when you run cold and warm audiences in the same CBO campaign.

Meta's job is to get you results.

You're using highest volume, a ROAS goal, or cost cap.

The algorithm scans your ad sets and finds the one most likely to deliver.

Warm audiences win. They always win.

Existing customers and recent visitors convert faster and cheaper than cold audiences.

Meta calls that optimization. It thinks its job is done.

What happens: your remarketing audiences absorbs most of the budget.

Frequency climbs fast. Cold audiences get almost nothing.

ROAS looks strong for a week or two, then drops as you exhaust the warm pool.

New customer acquisition quietly stops.

You don’t reach enough new customers which slows down growth.

You didn't ask Meta to do this.

But you gave it a CBO campaign and a results objective, and it made the logical choice.

The exception

If you run a small e-commerce shop with up to 5 products, an ABO campaign with one cold ad set and one warm ad set can work.

Manual control. You keep the split without building a separate campaign.

But most of my e-commerce clients have hundreds of products.

That's why I use a separate Catalog Sales campaign for remarketing, and I'll show you how I set it up.

How to set up an efficient catalog remarketing campaign for small budgets and larger ecommerce

Catalog Sales objective, product datafeed connected.

I use Advantage+ with audience controls.

This is the part most people misread.

Advantage+ here doesn't mean "no targeting."

Audience controls let you define who Meta should prioritize: your warm audiences, specific age ranges, past buyers, website visitors.

Meta starts there.

It may go broader when it finds better signals outside the warm pool.

Not always, but sometimes.

That's the trade-off.

You set the intent. Meta decides the path.

Audience controls: include existing customers and engaged audience.

No interest or behavior stacking. The warm pool does the segmentation.

Creative: trust frame on every product

The frame does what a static ad would take 5 seconds to communicate: social proof, a trust signal, a promo flag.

It works because the product is already the creative.

The frame adds the reason to buy.

For what to put on those frames and when to use them, read DAS #119.

I covered the full breakdown there.

Frame refresh: every 2 to 3 weeks.

During Black Friday or promo periods, I swap in a promotion-specific frame.

For accounts spending 1 000 to 5 000 EUR per month, I allocate up to 30% to remarketing.

Not a rule. A ceiling.

The actual number depends on your warm audience size.

If your site gets 200 visitors a month, you don't need 1500 EUR to remarket to them. You'll hit high frequency fast and burn through your warm pool in a week.

A rough check: frequency above 4 in the first week is a signal you're overspending for the audience size.

Pull the budget back.

This is the actual setup I run for one of my clients.

A direct-to-consumer brand, ~2500 EUR total monthly budget.

Remarketing gets 25 EUR/day, roughly 25% of total spend.

Goal: Sales.

Campaign: Catalog Sales, product datafeed connected.

Advantage+ with audience controls.

Attribution: 7-day click, 1-day view.

Audience controls:

Bid strategy: Highest volume.

At this spend level, bid caps restrict delivery without improving results.

Creative: Catalog ads with a custom trust frame.

This setup has been stable for months.

I don't merge remarketing back into prospecting when accounts grow.

The argument for consolidating is real. Fewer campaigns, simpler structure, let Meta optimize across the whole funnel.

But once you've built the audience controls and the catalog setup, the maintenance cost is low.

And you get something you can't get when everything is combined:

a clear read on what your warm audience spend is actually returning.

Keep the structures separate. Keep the budgets visible.

It needs more manual work.

But it means I have more control over my marketing funnel and audiences.

That's what I've found most effective, from 1000 EUR accounts to 5000 EUR accounts.

P.S. One day this may change. We might just drop a URL and Meta handles everything else. Until then, this is the structure I'd use.

TLDR