A testing environment that's been trickling in ad sets all year without a clean-up often looks fine in October, and then falls apart the moment BFCM spend hits it. You get too many ad sets competing in the same auction, overlapping broad audiences, and frequency climbing past healthy levels before peak.

<aside> ⭐

So, tidy up what's already there, then continue to build on top of it.

</aside>

What Actually Shifts Structurally

A few things move predictably once a BFCM offer goes live and the account needs to be shaped around them:

• Retargeting weight increases sharply for any brand running a live offer - spend naturally shifts toward warmer, more aware audiences during the sale window.
• CPMs rise as spend and competition increase across the board, which means every asset in the account needs to earn its place, not just add to volume.
• Saturation increases - audiences get hit with more ads than usual during this period, so creative that stands out visually matters more, not less.

[insert examples of creatives that stand out]

• BAU budgets keep scaling alongside the BFCM push. High-performing always-on campaigns don't get paused to make room for the sale - they keep running and scaling in parallel.

<aside> ⭐

None of this requires new campaign types or a different account architecture.

</aside>

It requires the existing structure to be more flexible (more weight here, tighter creative there) while staying fundamentally the same shape.

Keep a Dedicated BFCM vs. BAU Split

The single most important structural decision is keeping BFCM and always-on activity clearly split, rather than letting the sale swallow the whole account.

Push everything into most-aware, offer-led territory for the full period, and you're left with a hard fatigue cliff the moment the sale ends - no warm evergreen activity left running underneath it to catch the account when BFCM stops.

<aside> ⭐

The fix is a considered split: proven BAU assets keep running throughout BFCM, not just before and after it, while a dedicated BFCM layer sits on top for the sale window itself.

</aside>

That way the account exits BFCM in the same healthy shape it went in, rather than needing to be rebuilt from scratch in December.

Decide the Net-New vs. Iteration Split

Resourcing decisions are structural too, and they need to be made ahead of the sale.

• Put 70-80% of net-new effort into frameworks that are already proven - current top performers, or anything that's performed consistently across the full year, not just last BFCM.