Black Friday is November 27, and right about now a lot of store owners open Ads Manager, see a week of worse numbers, and reach for the pause button. Cost per purchase is up. The return on spend has slipped below where it sat in October. The obvious move seems to be to stop spending now, save the money, and turn everything back on when the sale starts.
For a store that is running a Black Friday sale, that move is almost always wrong. The numbers you are looking at this week are not lying, exactly, but they are incomplete in a specific and predictable way. Many of the people clicking your ads right now are gift shoppers who will not buy until your sale opens, and the default report in Ads Manager is built so that a lot of those sales never show up in it at all. Pause now, and you stop paying for the clicks that were going to turn into your best week of the year.
Here is how to see what your Facebook ads are actually doing, what pausing really costs you, and how to tell the difference between a slow week and a campaign that genuinely deserves to be cut.
Gift Shoppers Click Your Facebook Ad Now and Buy When Your Sale Starts
Picture a small online candle shop that sells gift sets for around $60. In the second week of November, someone sees its ad, clicks through, looks at three sets, and leaves. She is not losing interest. She is making a list. She knows Black Friday is two weeks away, she suspects the shop will run a sale, and she will come back when it does.
Multiply her by a few hundred and you have the normal shape of November for any store that sells things people give as gifts. Browsing goes up. Buying gets pushed to the sale. The clicks happen now and the purchases happen later.
That gap would not matter if Ads Manager waited for the purchases to arrive. It does not. Two settings decide what you see:
The attribution window. By default, every new ad set uses what Meta calls "7-day click or 1-day view." A purchase counts toward your ad only if it happens within 7 days of someone clicking it, or within 1 day of someone seeing it. Your candle shopper who clicked on November 14 and buys on November 27 bought 13 days later. In the default report, her purchase never shows up as a result of that ad.
The reporting date. Ads Manager files a purchase under the day the ad was shown, not the day the purchase happened. Meta's own help pages give the example plainly: if someone clicks an ad on Monday, Ads Manager reports on the impression that happened on Monday. So when a shopper who clicked last Tuesday buys this Sunday, the sale lands on last Tuesday's row. Every recent day in your report is still filling in, and the last few days always look the worst because their buyers have not finished buying yet.
Put those two together and you get the November trap. The most recent week looks weak because its purchases have not happened yet, and the purchases that will happen on Black Friday from clicks before about November 20 fall outside the default 7-day window entirely.
None of this means costs are not really rising. They are. Advertising space gets more expensive every fourth quarter as bigger retailers flood the auction, and we covered why in Meta Ads Get More Expensive in Q4. The point is that two different things are making November look bad at the same time: a real price increase you have to plan around, and a reporting gap that makes your results look worse than they are. Owners who pause in mid-November are usually reacting to both and blaming only the ads.
Meta Tracks Sales for 28 Days After a Click. Your Default Report Stops at Day 7.
You do not have to guess how many sales are hiding outside the default window. Meta will show you, but only if you ask. The tool is called Compare attribution settings, and it takes about two minutes to set up:
- Open Ads Manager and go to your campaigns.
- Click the Columns dropdown (it usually says Columns: Performance).
- Scroll down and choose Compare attribution settings.
- Tick 1-day click, 7-day click and 28-day click.
- Choose All conversions for the conversion count, then click Apply.
Your table now shows purchases and purchase value side by side under each window. The 7-day column is roughly what you have been looking at. The 28-day column adds every purchase that happened between day 8 and day 28 after a click. The gap between those two columns is the late-buying behavior your default report was leaving out. One caveat Meta spells out: if a campaign has been live for less than 28 days, the 28-day column only covers the time it has been running, so a brand-new campaign will understate the gap.
Here is what that can look like for the candle shop. These are example numbers, not a real account, but the pattern is the one to look for:
Under the default view, the shop is spending $1.00 to bring back $1.80 in sales. At its margins, that is below break-even, and the instinct to pause looks justified. Under the 28-day view, the same spend brought back $2.30, which clears the line. Same ads, same spend, same customers. The only thing that changed is how long the report waited.
How big is the gap in a real store? It varies, which is exactly why you should look at your own. Common Thread Collective, an agency that runs ads for large online brands, says that across its own accounts roughly 80% of sales value lands inside the 7-day window and about 20% arrives between day 8 and day 28, and that the late share grows by 5 to 15 points in the weeks before Black Friday. That is one agency's rule of thumb, not a Meta figure, and a small store selling $60 gifts may look different from a large brand. But it tells you the gap is not a rounding error.
Two more checks keep you honest:
Look at last November, too. If your account ran ads in the weeks before last Black Friday, set the date range to November 1 through 30 of last year and compare the 7-day and 28-day columns. That is the closest thing you have to a forecast of how your customers behave this month.
Count against your store, not just Meta. Meta's columns tell you what Meta can match to its own ads. Your Shopify orders tell you what actually happened. If your total store sales are holding up while Meta's 7-day numbers sag, the sales are still arriving; Meta is just not getting credit for them inside its default window. If you have never compared the two, our piece on Google Analytics missing holiday orders walks through how to do it in a way that applies here as well. And if you would rather see the purchases on the day they happened instead of the day of the click, Meta's Ads Reporting tool lets you switch the report to time of conversion. Neither view is wrong. They answer different questions.
Pausing Your Facebook Ads Sends Warm Shoppers to the Stores Still Running Theirs
Say you pause anyway. The money you stop spending feels like money saved. It is not, because a pause in mid-November costs you three things that are hard to buy back.
You disappear right before the purchase. The shopper who clicked your ad last week is still deciding. Over the next two weeks she will see ads from every store that sells something like yours. If you have gone quiet, the stores that kept running are the ones in front of her when she finally reaches for her card. You paid to start that sale; someone else gets to finish it.
Your retargeting pool stops growing. Every person who visits your site this month is a person you can show a Black Friday ad to later. That is the cheapest, warmest audience you will have all year. If you stop the ads that bring them in, the pool you will be selling to on November 27 is smaller than it would have been. We wrote about building that audience early in Black Friday Is Won in September, Not November; a pause in mid-November undoes part of that work.
You send your ads back to school on the worst possible day. Meta's delivery system spends the first stretch of every ad set in what it calls the learning phase, exploring who responds before it settles down. Meta says an ad set usually leaves learning after about 50 results in the week after its last significant edit, and it publishes a list of what counts as one. Pausing an ad set for 7 days or longer is on that list: the ad set re-enters learning when you turn it back on. Pause on November 14, unpause on Black Friday morning, and that is 13 days. Your ads would spend the most important day of your year relearning who your customers are, with less stable delivery and a higher cost per sale.
One Bad Day of Facebook Ads Tells You Nothing. Your Break-Even Number Says When to Cut.
None of this means you should never cut. It means you should cut on a number you chose in advance, not on a feeling you had on a Tuesday.
Start with your break-even return. This is the return on ad spend at which an order exactly pays for the ads it took to get it. For the candle shop, a $60 order leaves about $30 after the product, packaging, shipping and payment fees. Divide the order value by what's left, $60 by $30, and the break-even return is 2.0. Below that, every Meta sale loses money; above it, every sale makes some. If you have not worked yours out, Your Ads Report a 3x Return. At a 30% Margin, Every One of Those Orders Lost Money. shows how, including the costs most calculators leave out.
Judge the 28-day column, not the 7-day one. During November, the 7-day number will understate almost every gift-driven store. Measure your results against break-even using the 28-day click column.
Judge a window that has had time to fill in. Remember that Ads Manager files sales under the click date. Yesterday's row is mostly empty, and last week's row is still filling. Look at a 14-day stretch that ended at least a week ago, and watch how it changes over the next few days. If its 28-day return keeps climbing toward or past your break-even line, your ads are working and the recent days are simply unfinished.
Then decide, using one of three answers instead of two:
- Hold when the 28-day return is at or above break-even and you have a Black Friday sale planned. This is the common case, and the one most owners talk themselves out of.
- Trim when the 28-day return sits a little under break-even. Lower the budget in modest steps rather than turning it off. Meta says a budget change may or may not restart learning depending on its size: going from $100 a day to $101 is unlikely to, going from $100 to $1,000 may. Small trims keep the ads running and the retargeting pool growing.
- Cut when the 28-day return stays well below break-even even with late sales counted, when you are not running a sale and nothing you sell makes sense as a gift, or when a product simply cannot make money at fourth-quarter prices. Cut that product's ads, not the whole account.
One more reason to cut has nothing to do with seasonality: something is actually broken. A rejected ad, a declined payment card, a tracking change or worn-out creative can all look like a November slump. Before you blame the calendar, run through the checks in Why Did My Facebook Ads Stop Working?. If one of those is the cause, fix it. Holding spend on a broken campaign is not patience.
If You Raise Your Facebook Budget When You Wake Up on Black Friday, Your Sale Has Been Running for Hours.
Holding through mid-November gets you to Black Friday with your ads warmed up and your audience full. Black Friday week itself rewards a little planning, because the timing works differently from the rest of the year.
Set your budgets the night before, in the right time zone. Meta's day starts at midnight in your ad account's time zone, which is not necessarily where you live, especially if someone else set the account up. If your sale opens at midnight and you raise your budget when you wake up, your ads spent the first hours of the sale at their old level. Raise budgets on Thanksgiving evening for Friday, and on Sunday evening for Cyber Monday. Check what time zone your account runs on before you do.
Get your sale ads approved before the week starts. Meta reviews new ads before they run, and adding a new ad to an ad set is one of the significant edits that restarts learning. Upload your Black Friday creative by Monday, November 23 at the latest, so it has cleared review and settled in before the sale. Adding frames or a sale banner to ads that are already working is often faster and safer than building everything from scratch.
Give sale products their own campaign only when the math changes. If your sale is 25% off everything, the return you need from your ads changes, because each order now leaves you less money. Products whose margin changes the same way can stay together in one campaign. Split out a separate campaign only when a group of products needs a genuinely different target, for example a deeper discount on one line, or a separate campaign aimed at past customers. Simpler structures give Meta more data per ad set, which matters in the week everything moves fastest.
Bring the budget back down before Tuesday. The demand that justifies a big Black Friday and Cyber Monday budget is gone by Tuesday, December 1. Set a reminder to bring budgets back to normal on Monday night. Leaving the weekend budget running into an ordinary Tuesday is one of the easiest ways to give back some of what the weekend earned.
If you also run Google Ads, the advice there is different, because Google's systems handle the season on their own; we laid that out in Should You Change Your Google Ads for Black Friday?. On Meta, the job this month is mostly patience: read the right column, set your break-even line before you look, and keep paying for the clicks that are going to buy on the 27th.
We help online stores plan this every year, from the break-even number to the budget calendar for the week itself. If you would rather spend November running your shop than watching attribution columns, that is the kind of thing we can take off your plate. If you would rather do it yourself, open Compare attribution settings today. It will tell you more about your November than any single day of results ever could.
