Why Your Black Friday Ads Stop Working in Week Two

Most peak performance decline gets blamed on creative. The more common cause is frequency: the same people seeing the same ads more often, because budget grew and the audience did not.

The decline usually shows up in the second week, and the settings that control it are structural, which means they lock this week ahead of 27 November. This post covers what ad frequency does to performance, where overlap hides, and the exclusions worth setting before the freeze.

Why does frequency rise faster than spend?

Because the audience is finite and the budget is not.

Doubling spend against a pool that has not grown does not reach twice as many people. It reaches roughly the same people twice as often. The delivery system fills the budget from the cheapest available impressions, and the cheapest impressions are the people it has already shown your ad to.

Peak compounds this. Every brand raises budget in the same fortnight, so the cost of reaching someone new rises while the cost of reaching someone familiar stays comparatively low, which pushes delivery further toward repetition.

What does rising frequency look like in the numbers?

Cost per click rises and click through rate falls first. Conversion rate moves later, usually about a week later.

That lag is why frequency gets misdiagnosed. By the time conversion rate drops, the account has been paying for the decline for days, and the obvious reading is that the creative stopped working. Often the creative is fine and the audience is saturated.

Check frequency alongside those two metrics daily during peak rather than weekly. It is the variable that explains the other two.

Where does overlap hide?

Between campaigns that look separate on paper.

Three common cases. A lookalike audience that substantially contains your existing customers. A broad prospecting campaign that includes everyone your retargeting campaigns are also bidding on. And several retargeting windows, at 7, 14 and 30 days, running simultaneously without exclusions, so a recent visitor sits in all three.

The result is your campaigns competing against each other in the same auction, which raises your own costs. The account looks like it has five audiences and behaves like it has two.

What exclusions should be in place?

Four, and all of them are structural changes that belong before the freeze.

Recent purchasers excluded from acquisition campaigns, with the window matched to your repurchase cycle rather than set at a default. Existing customers excluded from prospecting. Shorter retargeting windows excluded from longer ones, so each tier reaches a distinct group. And anyone mid way through a high intent flow excluded from broad promotional messaging.

Purchaser exclusion is the one most often missed during peak, and it is the most expensive, because the person who just bought is the cheapest person to show an ad to and the least likely to buy again that week.

Should you cap frequency?

Set a ceiling you would be uncomfortable passing, then watch it rather than enforcing it rigidly.

Hard caps constrain delivery in a window where you want the system spending, so they can cost volume on the days that matter. The more useful approach during peak is treating frequency as a diagnostic: if it climbs while click through rate falls, the answer is more distinct creative or a wider audience rather than a tighter cap.

The exception is retargeting, where a cap earns its keep. A warm audience seeing the same offer twelve times in a weekend converts no better and complains more.

What actually fixes saturation?

More genuinely distinct creative, or a bigger pool. Those are the only two real answers.

Distinct matters more than numerous. Fifteen versions of the same concept read as one concept to the delivery system, so a large bank of near duplicates behaves like a small bank while costing more to produce. Different angles, formats and storylines are what reach fresh audiences, which is covered in UGC ads at peak.

Widening the pool is largely out of time now, since reach bought this late barely enters the audience before the weekend ends. That window is covered in auditing your retargeting pool.

What should you avoid doing mid peak?

Narrowing audiences to improve the reported numbers, and restructuring to solve a frequency problem.

Narrowing makes frequency worse by shrinking the denominator, even though the efficiency figure improves briefly. Our practice is loosening demographic guardrails from early October through to mid January rather than tightening them.

Restructuring mid peak resets learning during the most expensive days of the year. Scale by no more than 30% in any seven day period and change creative rather than structure, which is covered in what breaks if spend doubles.

The account operating rules for the whole season sit in the Q4 Profit Playbook, free and written for founder led DTC brands on Shopify at $5M to $30M. Webtopia runs peak media for founder led brands as an ecommerce paid media agency.

Frequently asked questions

What is a good ad frequency during Black Friday?

There is no universal number. Watch it against click through rate and cost per click instead: frequency rising while click through rate falls is the signal that matters.

Why does performance drop in the second week of a campaign?

Usually audience saturation rather than creative fatigue. Spend rose, the pool did not, so the same people are seeing the ads more often.

Should I exclude recent purchasers from peak campaigns?

Yes, from acquisition campaigns, with the window matched to your repurchase cycle. It is the most commonly missed and most expensive exclusion during peak.

Do frequency caps help at peak?

On retargeting, yes. On prospecting they can constrain delivery on the days you most want the system spending, so treat frequency as a diagnostic there instead.

Does more creative fix a frequency problem?

Only if it is genuinely distinct. Many versions of one concept read as a single concept and reach roughly one concept's worth of audience.