Brandpeneur
Abstract cool-toned grid of small test tiles with one highlighted, suggesting creative testing in a quiet month
Meta AdsQ1

January Is the Best Month to Test Meta Ads and the Worst to Judge Them

Joshua George|18 November 2026

A january meta ads strategy should have one job: find the creative that carries you through the next three months. It should not try to hit the ROAS you saw in November. January is a good month to test and a poor month to judge.

Auctions are usually calmer once the holiday rush ends, so a failed test costs less. Shoppers are also more careful with money, so the headline ROAS tells you less than usual about the health of the account. Plan the month around testing, judged on your break-even, and you will start February with a stronger account than most.

Why January suits testing

Through Q4, more brands compete for the same people, which pushes the price of reaching them up. The reasons are laid out in why CPMs are so high in Q4. When that pressure eases, the cost of showing a new ad to a few thousand people usually falls with it.

That matters because testing is a spend with an uncertain return. Most new ads lose. The cheaper it is to find out that an ad loses, the more concepts you can afford to try. A test that would eat a painful share of your margin at peak prices is a modest cost in a quiet month.

Prices do not always fall. Some categories get their own January demand, such as fitness, wellness and tidying products. Look at your own CPM in the last two weeks of December and in January last year before you plan around a drop.

Why January ROAS misleads

Three things bend the number at the same time, and they point in different directions.

  • Cautious buyers. After Christmas, many shoppers hold back. Conversion rates on cold traffic can dip even when your ads are fine.
  • A thin pool of ready buyers. Your best prospects bought in November or December, so there are fewer people about to purchase.
  • Tests drag the average. New ads spend before they have proved themselves. Blended ROAS falls while the testing runs, even when the work is right.

So a ROAS that looks worse than Q4 is not evidence that something broke. It also does not tell you whether you are making money. That depends on the margin of each order, which is why I start with the maths. The December playbook covers resetting that number once the sale ends, and the same rule applies here.

Set your January break-even first

Everything in the month hangs off two numbers: your max CPA and your break-even ROAS. Both come from contribution margin, which is what is left from each order after product cost, shipping, payment fees and returns.

Say you sell at £60 with a 40% contribution margin. That leaves £24 per order. Your max CPA is £24 and your break-even ROAS is 2.5. If January's average ROAS is 2.8, you are making a little on every order. If it is 2.2, you are losing money on each one, however good it looks next to the sale months.

Now split the budget into two parts, with a different rule for each:

PartPurposeJudge it on
Proven adsKeep the account profitableCost per purchase against max CPA
Test adsFind the next winnersWhether a concept earns a place in the proven group

Say your daily budget is £600. You might give £450 to proven ads and £150 to tests. The test share is a decision, not a percentage anyone can prescribe. Size it so that if every test fails, the month is still acceptable to you.

How to run creative tests in January

Most weak tests fail for the same reason: too many ads, too little spend on each, and no way to tell the results apart. Keep it simple.

  1. Test concepts, not tweaks. A different angle, a different person speaking, a different problem named in the first line. Changing the button colour or one word teaches you very little.
  2. Limit how many run at once. Work out what an ad needs to spend to produce a handful of purchases at your CPA. If you cannot fund that for an ad, you have too many ads.
  3. Wait for purchases before you judge. Early clicks and hold rates can flag a clear dud. They cannot crown a winner.
  4. Write the kill line before you launch. For example, an ad that has spent two times your max CPA with no purchase gets paused. Deciding in advance stops you defending a favourite.
  5. Promote winners into the proven group. An ad that clears max CPA over enough spend moves across. Do not just raise its budget in place and hope.

Say your max CPA is £24. An ad that has spent £48 with no purchases hits the kill line. An ad that has spent £120 and produced five purchases is at £24 each, right on the line, and probably needs more spend before you decide. The ad that produced five purchases for £90 is at £18 and has earned its place.

Start the month with what you already know. Take your best ad from Q4 and make three genuinely different versions of its core idea. Winners usually come from understanding why an ad worked, not from random new ideas.

How much to spend in January

Work from the break-even, not from the previous month. The mistake I see most is carrying a Q4 budget into January because it feels like the account should keep growing. Demand is lower, so that spend often runs past the point where it pays.

The test is marginal return, which is the revenue from your last extra pound rather than the average across the account. Say you raise spend from £500 to £600 a day and revenue climbs by £200 a day. The extra £100 returned £2.00 per pound. If break-even is 2.5, that step lost money, even if the account average still shows 3.0. Hold the earlier level.

When you want to know where your own limit sits, work out the break-even figure with the break-even calculator, then read your last few budget changes against it.

What not to do in January

  • Do not copy Q4 targets into Q1. The offers, the buyers and the auction are different.
  • Do not reset the account. Rebuilding campaigns because ROAS dipped throws away learning that took months to build.
  • Do not pause everything. Brands that go dark lose the data and the audience signal, then pay to rebuild both.
  • Do not judge a test on three days. Wait for purchases, then apply the kill line you wrote.

What you want to have by February

By the end of the month you want a short list of ads that cleared max CPA, a clear idea of why they worked, and an account that stayed profitable on its proven group while the tests ran. That list is what you scale in spring, when demand returns and spend can rise without paying the price of an untested account.

If you want someone to check your numbers before you plan the month, the free profit audit works through contribution margin, max CPA and marginal return for your store. The growth service page explains how the work is run for brands that want it done for them.

Running a Shopify store?

See what a team that actually communicates looks like. We will audit your ad account, calculate your real numbers, and show you exactly where the profit is hiding.

Book Your Free Profit Audit
← Back to all posts