Black Friday is won in October

Every October I have the same conversation with founders.

"We're going big on Black Friday this year."

Then November arrives, and the same problems show up. I've seen them from both sides of the table, in-house and agency, and they come round every year like clockwork.

The four mistakes I see every year

1. Doubling the budget overnight

An account that has never spent more than £1,500 a day suddenly gets £5,000. Meta has no history at that level, the learning resets, and costs jump at exactly the moment they are already highest. Spend needs to build towards the sale, not leap on the day.

2. Deciding the offer the week before

Usually it ends up as 20% off, because that's what happened last year. Nobody has checked the margin at that discount once ad costs rise, and nobody has looked at what competitors are about to run. The offer is the single biggest lever of the whole period, and it often gets the least thought.

3. Last month's ads with a "SALE" banner on top

Your audience has already seen those ads. A red banner doesn't make them new. During BFCM people are comparing offers across a dozen brands, so your creative has to explain quickly why yours is worth it.

4. Tracking quietly breaking

A new landing page, a discount app, a checkout change. Something small breaks the purchase event, and nobody notices until December, by which point you have been optimising on bad data through the most expensive week of the year.

None of this is a budget problem. It's a timing problem.

Why the timing matters so much

Ad costs climb sharply through late November, because every brand is bidding for the same people at once. Triple Whale's data from BFCM 2025 showed Meta took around two thirds of all ad spend over the period, and the cost per purchase still came in slightly higher than the year before.

That changes how you should use the weeks beforehand. October and early November are your chance to test new angles and offers while costs are still normal. Sale week is for spending hard on the things you already know work. If you're still testing during BFCM itself, you're paying peak prices to learn.

What good looks like

This is the rough timeline I work to. If you're reading this in October, you're not late, but the window is closing.

• Early OctoberHealth check the accountCheck tracking end to end, tidy up campaign structure and make sure your best performers can take more budget.
• Mid OctoberLock the offerCheck the margin at your planned discount with realistic ad costs. Look at what competitors ran last year and what they are teasing now.
• Late OctoberBrief and produce BFCM creativeNew concepts built for the sale, not reskinned ads. Test the angles early at normal costs.
• Early NovemberBuild demandGrow your email and SMS lists with early access, warm up audiences and start building spend gradually.
• BFCM weekScale what's provenPush budget behind proven ads, refresh before fatigue sets in and keep a close eye on frequency.
• DecemberWind down slowlyDon't cut spend off a cliff. Bring it down gradually and keep remarketing to the people who nearly bought.

Think beyond a percentage off

A flat discount is the easiest offer to copy and the hardest to stand out with. Depending on your margins, these often work harder:

• Bundles that lift order value while protecting margin
• A gift with purchase above a spend threshold
• Tiered offers, where the more you spend, the more you save
• Early access for subscribers, which also grows your list ahead of the sale

A quick readiness checklist

If you can tick all of these by early November, you're in a better position than most brands I speak to.

• Your purchase tracking has been tested end to end in the last two weeks
• Your offer is locked and you know your margin at realistic ad costs
• You have new creative made for the sale, not reskinned ads
• You know which ads can take more budget without breaking
• You have a plan for how spend builds up, peaks and winds down
• Your site, checkout and stock can handle the extra demand