18 Common BFCM Mistakes and How to Avoid Them
· PPC

Last Q4 (2025), the median account we analyzed spent 3.8 times its normal daily budget on its biggest day of Black Friday weekend, with one account reaching 67 times. Whether all that extra spend turns into profit gets decided weeks earlier in October.
The mistakes below are the ones we see most often in accounts heading into the peak, across Meta, Google and creative. Almost all of them take root in the weeks before Black Friday, while the audience, the account structure and the creative are still being decided. With October still ahead of you, there is time to fix most of them before the peak.
The earliest warning sign for Meta is CPMr, or Cost Per Mille Reached, the framework our Director of Paid Social Phil Kiel popularized for catching saturation before ROAS reacts. You calculate it by multiplying CPM by frequency, which gives you the cost of reaching 1,000 unique people. Meta reports it in Ads Manager as "Cost per 1,000 Accounts Centre accounts reached." When CPMr rises week over week, your budget is cycling through the same people, a shift that shows up weeks before CTR or ROAS move.
Meta
Arriving at Black Friday already saturated
The most common thing we find going into Q4 is an account that is already saturated before the Black Friday budget arrives. Spend holds steady through September while reach quietly shrinks, frequency creeps up and CPMr climbs, with ROAS flat enough that nobody looks twice.
Take an account spending $75k a week that reaches 1.9M people in August and 1.1M by the end of September on the same budget. Its CPMr has risen by around 75% without a single bad week showing in ROAS. If Black Friday budget then gets layered onto that worn-out audience, you pay peak auction prices to reach people who have already seen your ads several times, which is why exclusions, consolidation and fresh creative all need to happen in October.
[https://x.com/PhilKiel/status/2029701498268176617]
Treating every CPMr rise as the same problem
A rising CPMr can come from 2 very different places, with a different fix for each. When frequency does most of the climbing while CPM holds, the budget is cycling through the same people, a problem exclusions and consolidation will fix. If CPM is doing the climbing while frequency holds or falls, the pressure is coming from the auction. In that case, check placement mix, bid caps and where the extra budget went before you touch targeting at all.
We see both in the run-up to Black Friday, sometimes in accounts on almost identical budgets, with a headline number that looks the same in each case.
[https://x.com/PhilKiel/status/2025315598541140468]
Budgeting for the peak day instead of the peak month
Across the 12 accounts in our fleet with enough Q4 2025 history to compare, the median account's biggest single day ran at 3.8 times a normal August 2025 day. Its heaviest 30 consecutive days ran at 1.8 times August. One account held 26 times its August spend for a full month.
Card limits and billing thresholds get consumed by that accumulation over weeks. When they hit on Cyber Monday, ads stop serving on the most valuable day of the year. Plan payment capacity against your projected peak month, then raise account spending limits and card limits before November.
Letting retargeting take the credit
When budgets jump in Black Friday week, spend usually grows much faster than reach, so a growing share of the budget lands on people already in the funnel.
Warm ad sets serving discount messaging to people halfway to checkout will usually top the ROAS column, while the prospecting ads doing the acquisition work get cut. Holding exclusions through the peak, with warm and cold ad sets judged against separate benchmarks, keeps the read honest.
[https://x.com/PhilKiel/status/2019806756356931929]
Restructuring the account in November
Consolidation is the right call for a saturated account, but timing decides whether it pays off. Ad sets rebuilt 2 weeks before Black Friday are still finding their footing when the auction is at its most crowded. Structural changes belong in October, leaving November for scaling what is already stable.
Switching prospecting off after Cyber Monday
CPMs usually drop back quickly once Cyber Monday is over, which makes the first half of December some of the lowest-cost reach of the quarter. Many accounts pull prospecting back hard at exactly that point. Keeping it running into December builds the audience you will convert in January, instead of starting the new year from a pool you have already exhausted.
Giving Google the credit for demand Meta built
In Black Friday week last year, the Automotive account's Google campaigns returned a 6.5 ROAS on $131k of spend, up from 5.9 the week before. Over those same 7 days, Meta spend rose 5.5 times while Meta reach grew 3.4 times, matching the 3.4x lift in Google conversions.
Search captures Black Friday demand far more than it creates it. Judging channels on platform ROAS during the peak overstates Google and understates whatever filled the search bar in the first place. Blended numbers such as MER across both channels give you a far more honest read of the week.
Leaving peak budgets running after Cyber Monday
In the second week of December, the same account was still spending 21% more on Google than before the peak, while ROAS had dropped to 3.2 against 5.1 in October. Budgets and targets raised for the peak tend to stay wherever they were left. Smart bidding then keeps spending into a conversion rate that has already fallen away.
Put the reset date in the calendar before Black Friday, so December does not quietly eat the margin November made.
Letting smart bidding guess at the conversion rate spike
Smart bidding learns from recent history, so it lags behind the jump in conversion rate during Black Friday week and then overcorrects once the peak ends. Seasonality adjustments exist for exactly this kind of short event. Set them for the specific days instead of leaving the algorithm to catch up while the most valuable traffic of the year moves through the auction.
Letting Performance Max feed on branded search
When branded search volume climbs in Black Friday week, Performance Max will happily pick that traffic up and report it as its own performance. Without brand exclusions in place, PMax ROAS looks strongest exactly when it contributes least, which then justifies more budget flowing into it for next year's peak. Apply brand exclusion lists before November, then use the channel performance view during the week to see where conversions actually come from.
Running out of budget on the campaigns that matter
Your highest-intent search campaigns are usually the first to hit their daily cap in Black Friday week, often by mid-afternoon. Check search lost impression share due to budget daily through the peak, because every hour a capped campaign sits dark hands that intent to a competitor.
Skipping the search terms review in November
Black Friday queries pull in a wave of competitor terms, deal-hunter searches and loosely related products, especially on broad match. Review search terms and build negatives in the weeks before the peak, so the extra budget stays pointed at people searching for what you sell.
Submitting promotions too late to clear review
Merchant Center promotions and sale price updates both go through review, with the queue at its longest in the run-up to Black Friday. Submitting them at least 2 weeks early means the offer shows on Shopping ads on the day, instead of the product appearing at full price while the discount sits in pending.
Creative
Launching new creative in Black Friday week
Meta's Andromeda retrieval system uses creative as a targeting signal, so every new ad needs time to find its audience. Black Friday week is the most expensive moment of the year to start that process. The accounts that handle peak season best have core creative live in October, then build Black Friday versions as offer overlays on ads that have already earned their reach.
Defaulting to percentage-off without testing other offers
Across our accounts over the last 90 days, free gift with purchase beat the account's own baseline in all 4 accounts that tested it, with a median lift of 27%. Free shipping won in 5 of 7 accounts. "Up to 50% off" won in 3 of 5, with a median lift of 9%.
These samples are small, so treat this as a reason to put value-add offers into test alongside the discount. Most accounts never test anything beyond the percentage.
Assuming urgency works everywhere
Urgency is the strongest message type in our data, beating baseline in 27 of the 40 accounts that tested it. "Limited time offer" on its own won in 12 of 17 accounts. Where urgency lost, it lost heavily, with 2 accounts down 49% and 60% against their own baselines. Test it in your account before building an entire Black Friday library around a countdown timer.
Making 20 versions of the same offer ad
Black Friday production usually means one discount concept resized and recolored until there are 20 ads that Meta treats as one. Running 10 distinct concepts with the same offer reaches more people than 20 variations of a single idea, keeping CPMr lower through the week when every account is competing for the same attention.
[https://x.com/PhilKiel/status/2026310417103126532]
Pausing proven winners to make room for Black Friday ads
The evergreen ads that carried the account all year, often a Yapper (a direct-to-camera talking-head ad) or a creator-led Partnership Ad, tend to get paused in favor of brand-new sale creative. Those ads have already found their audience, which keeps CPMr down while the new offer ads learn. Run the offer as an overlay or first-frame change on a proven winner, so that reach keeps working through the most expensive week of the year.
Where to start this week
The common thread across all 18 mistakes is timing. Most of the damage in Q4 gets locked in during October, through audiences left to saturate, structures rebuilt too late and creative launched into the most expensive auction of the year.
Pull CPMr for the last 8 weeks and split it into CPM and frequency. If frequency is doing most of the climbing, you have until the end of October to widen the audience before it becomes an expensive problem to fix mid-peak. For a full walkthrough of the math, read Phil Kiel's guide tocalculating CPMr. If you want our team to run this diagnostic on your account,book a call.
Methodology: Data from Taikun-managed accounts, anonymized and pulled through Data Druid. Meta and Google weekly data covers August to September 2026 and Q4 2025. The fleet spend benchmark covers the 12 accounts with enough history, measured against August 2025 daily spend. Offer and message benchmarks cover 78 accounts over the last 90 days, measured against each account's own baseline.