How to test turning off Meta ads, step by step

Pick some regions that move like the rest and take Meta out of them in every ad set. Lower budgets so the other regions get no extra money. Then compare Shopify sales in both groups for longer than your slowest buyers take. If the regions without ads barely dip, Meta was mostly buying sales that were coming anyway.

You need Ads Manager, Ads Reporting, your Shopify admin and a spreadsheet, plus a few weeks of patience. The test asks one thing: what happens to sales where Meta goes dark? Write the answer rule down first, so the result cannot talk you into anything.

Step by step

1. Set the pass mark first. Divide 1 by your gross margin. The result is the sales each euro of Meta spend must cause to pay for itself. Keep Meta only if the dark regions lose more sales than the spend saved times that number. Menu path: Shopify admin > Analytics > Reports > Category filter > Profit Margin, then read Gross margin.
2. Find your regions in Shopify. The Total sales by billing location report shows sales by the country or region of each order's billing address. Shortlist regions with no shop, event or local promotion of their own. Menu path: Shopify admin > Analytics > Reports > Category filter > Sales > Total sales by billing location.
3. Build a weekly baseline. Group the report by week over the past few months, test and control regions apart. Good test regions moved in step with the control regions before anything changed. Menu path: in the report's configuration panel, Dimensions > add icon > Week.
4. Decide what goes dark. Switching off every campaign in the dark regions answers whether Meta pays at all. To test one campaign, rank them first in Meta's Incremental column, which counts only conversions its models consider incremental. Meta advises comparing campaigns within that column, not against standard results. Menu path: Ads Manager > Columns: Performance > Compare attribution models > Incremental > Apply.
5. Take the dark regions out of every ad set. Edit each ad set's Locations to list the control regions only. For a city test, use Exclude cities instead. Untick Reach more people likely to respond to your ads, or Meta may reach beyond your list. Menu path: Ads Manager > edit the ad set > Audience > Locations.
6. Cut each budget by the dark regions' share. Ads Reporting shows last month's spend by region, so you can see that share. If the dark regions took a fifth of the spend, take a fifth off each budget. Otherwise the same money usually chases fewer people in the control regions and lifts their sales. Menu path: Ads Manager > hover over the ad set or campaign > Edit > change the budget > Publish.
7. Restart the clock after the edit. Meta counts any targeting change as a significant edit, so each edited ad set relearns. Count the test from that day, and make no other edits until it ends. Menu path: Ads Manager > Ad sets > Delivery column, adding Last significant edit through Columns.
8. Mark the test on your sales chart. Shopify's Ad spend change annotation covers a paid campaign that starts or stops. Add one that spans the whole test. Menu path: Shopify admin > Analytics > Reports > Total sales over time > Annotations > add icon > Start and End > Category.
9. Check that the dark regions really went dark. Ads Reporting breaks delivery down by region, so Amount spent there should fall close to zero. It does not break website purchases down by region, so read sales in Shopify. Menu path: Ads Reporting > breakdowns > Geography > Region, with Amount spent as the metric.
10. Run past your slowest buyers, then compare. Set each group's sales during the test against its own baseline. Scale the dark regions' baseline by the control regions' change; the gap to what they actually sold is what Meta was adding. Menu path: Shopify admin > Analytics > Reports > Total sales by billing location > Compare to > Comparison to past.

A worked example

For illustration, take a shop with €40,000 a week of sales in its control regions and €10,000 in its dark regions.

For illustration, the control regions slipped 5%, so the dark regions would have sold about €9,500 a week with the ads. For illustration, they sold €9,000, so going dark cost about €500 of sales a week.

Now the pass mark. Take it from one store's anonymised GA4 export, 1 January 2024 to 21 August 2026. On its Break-even sheet, a 40% margin puts break-even ROAS at 2.5x, because 1 divided by 0.40 is 2.5. That row is arithmetic for any store at that margin, not that store's result.

For illustration, the €1,000 saved each week would have had to protect €2,500 of sales to pay for itself. For illustration, it protected about €500, a return of 0.5 on each euro against a line of 2.5. For illustration, at a 40% margin those €500 of sales left €200 of gross profit, so the ads in those regions lost €800 a week.

Ads Manager can show a healthy Purchase ROAS for the same ad sets all the while. It counts purchases that happened near an ad, needed or not.

The clock matters as much as the money. On that one store's Journeys sheet, journeys of 10 or more touches took 16.0 days to buy. If your slow buyers take that long, a two-week test ends before the last of them decide. Run it well past that point, so the slow buyers in both groups finish.

What to check when the report looks wrong

• The dark regions still show spend. Meta targets people by where they live or were recently. Someone from a control region can see your ad while visiting a dark one, so a trickle is normal. A steady stream is worth a look at the Reach more people box.
• The control regions jumped when the test began. The budgets probably kept their old size, so the same money went to fewer people. Lower them by the dark regions' share and start the clock again.
• Both groups fell together. That is the season, a stock-out or a skipped email, not Meta. It is why each group is judged against its own baseline.
• An edited ad set reads Learning limited. Narrower targeting can starve an ad set of results. Meta suggests combining similar ad sets and setting budgets large enough to get results.
• Shopify and Meta disagree about where buyers are. Shopify's report uses the billing address, while Meta targets where people live or were recently. A buyer who saw your ad on holiday still pays from home.
• The dip is real but small. That can be the answer: Meta added little there. Set it against the pass mark from step 1 before you call it noise.

What to do this week

1. Run Total sales by billing location for the past few months. Pass: a set of regions whose weekly sales moved with the rest. Fail: no region tracks the others, so use bigger regions or gather more history first.
2. Check whether Meta will run the split for you. Look for Conversion Lift under Experiments. Meta's guide asks for a campaign from the past year with $5,000 USD or more of spend and at least 500 conversions. Pass: you qualify, so let Meta build the groups. Fail: you do not, so the regions are your test.
3. Write the plan on one page. Regions, dates, budget cuts and the pass mark from step 1, agreed before any setting changes. Pass: it is written and dated. Fail: it gets written after the results arrive.

Check the homework. Your GA4 Attribution paths export already holds the evidence. Causality Engine reads that one file and shows what each channel caused next to what last-click gave it, in 1 to 2 minutes, for €99 once (excluding VAT), refundable within 30 days. Check the homework

Sources, 1 October 2026: Profit reports (Shopify Help Center); Sales reports (Shopify Help Center); Setting and comparing time ranges for your reports (Shopify Help Center); Annotations in your Shopify reports (Shopify Help Center); How to view results for incremental attribution in Meta Ads Manager (Meta Business Help Center); Use location targeting (Meta Business Help Center); Change your budget in Meta Ads Manager (Meta Business Help Center); Significant edits and learning phase (Meta Business Help Center); About the learning phase (Meta Business Help Center); About breakdowns, metrics and filtering in Meta Ads Reporting (Meta Business Help Center); About Conversion Lift (Meta Business Help Center).

Related answers

• All Meta ads questions
• Should I turn off my Meta ads?
• How to run a geo holdout test, step by step
• Measuring Meta ads incrementality, step by step
• Incrementality testing, explained