Our team strictly communicates with clients via @arlox.io email.

Back to JournalMost popular

Kill Winning Meta Campaigns Before the Cliff

We killed an 11-week campaign with target ROAS because frequency was climbing and new customer acquisition had stalled. The warm pool was exhausting. Two weeks of rebuilt momentum got us a lower cost per purchase than the winner we replaced.

A

Arlox Team·Aug 7, 2026·1 min read

A Meta ad campaign dashboard showing rising frequency and stalling new customer acquisition

We killed a campaign that was technically performing. The client thought we'd lost our minds.

ROAS above target. Spend efficient. Client was happy.

We killed it anyway.

The same creative had been running for 11 weeks. ROAS held, but frequency was climbing fast. The same people kept seeing the same ad. New customer acquisition had stalled. The campaign was retargeting the same warm pool on repeat and calling it performance.

It felt like it was working. It was running out of road.

We flagged it on a call. This campaign is on borrowed time. The moment that warm pool exhausts, we hit a cliff. Better to replace it now with fresh creative while we still have budget and momentum than scramble when the numbers fall.

The client wasn't convinced. Asked us to leave it running.

We pushed back. Showed the frequency data. Showed the drop in new customer acquisition rate. Showed the audience overlap.

They agreed. We killed it. Launched a new campaign with fresh creative and a cold audience strategy.

Took 2 weeks to build momentum. Then cost per purchase beat what the old campaign was delivering.

Knowing when to kill a winner is harder than finding one. At arlox.io it's one of the most important calls we make for D2C fashion brands on Meta.

If your Meta campaigns have been running the same creatives for months, something needs replacing. arlox.io.

- Arlox.io | Best Brand Scaling Agency for D2C Fashion

Key Takeaways
  • Target ROAS does not equal a healthy campaign.
  • Frequency climbing means your warm pool is exhausting.
  • New customer acquisition stalling means you are retargeting the same people.
  • Replace winning creatives before performance drops.
  • Accept two weeks of rebuilt momentum to get a lower cost per purchase.
The Short Answer

When should you kill a winning Meta campaign?

Arlox runs paid acquisition for D2C fashion brands doing $80K to $500K a month across India, the UK, the US and Nigeria. Your Meta ROAS can hold while your campaign dies. We killed an 11-week campaign with target ROAS because frequency was climbing and new customer acquisition stalled. The trade-off is two weeks of rebuilt momentum. We replaced the winner with fresh creative and a cold audience strategy. Cost per purchase beat the old campaign after two weeks. Kill your winners before the warm pool exhausts.

Keep reading

Broad Meta Targeting Is Waste With Extra Steps

1 min read

Advantage+ Automates Distribution. It Does Not Automate Strategy.

2 min read

Your Meta Ad Doesn't Have a Targeting Problem

1 min read

We Sent a Client Our Mistakes in a Spreadsheet. They Asked to Invest More.

3 min read

Your Bestseller Can Stall Your Meta ROAS

1 min read

Why We Paused Meta Ads 48 Hours Before a Major Sale

2 min read