Should I exclude the test regions or target the control regions only?

Either keeps brand ads out of the test regions. Google's geo guide targets the control regions only and sets Location options to Presence to prevent leakage. Whichever you pick, check the budget, because a capped campaign can push the money it saves into the regions that still run ads.

Can I test brand Shopping ads the same way?

Yes, but score them elsewhere. Google's paid and organic report counts text ads only, not Shopping ads. Fence Shopping and Performance Max with brand exclusions, then judge the test on Shopify sales by billing region rather than on any single ad report.

What if my regions sell very unevenly?

Pair regions that moved together in the baseline weeks, and judge each group against its own history, not against the other group's size. If one region carries most sales, a time-based design built for few regions fits better than a random split.

How to test branded search incrementality, step by step

Put every paid brand search in one campaign, record a baseline of brand clicks and sales, then switch brand ads off in some regions. Keep everything else still, count a cooldown, and compare total sales and paid plus organic clicks with the regions that kept their ads.

By Joris van Huët, Founder & CEOUpdated 2026-10-019 min read

Run the numbers for your store: the free holdout test planner.

Usually in nine steps. Fence your brand searches into one campaign, record a baseline of brand clicks and sales, then switch brand ads off in a set of regions. Keep everything else still, count a cooldown, and compare total sales and paid plus organic clicks with the regions that kept their ads.

You need Google Ads, Search Console, GA4 and Shopify, plus a run of weeks with no launch or sale. Steps 1 to 4 build the baseline, and steps 5 to 9 run and score the test. Skip the baseline and you end up comparing the test with your memory.

Step by step

Fence your brand searches into one campaign. A pause only tests something if one campaign holds all your paid brand clicks. Filter the search terms report for your name, then add your brand list to Performance Max as a brand exclusion. Path: Google Ads > Campaigns > Insights and reports > Search terms. Then: your Performance Max campaign > Settings > Additional settings > Brand exclusions.
Record organic brand clicks by week. Search Console's Branded filter keeps searches that include your brand name, domain or brand products. Read it by week, which Google suggests to even out the day-of-week effect. Path: Search Console > Performance > filter bar > Query > Branded.
Link Search Console to Google Ads. The paid and organic report then lists ad clicks, organic clicks and combined clicks for each search term. It only holds organic data from the day you link, so link before the baseline weeks. Path: Google Ads > Campaigns > Insights and reports > Report editor > Basic tab.
Pick regions, or pick weeks. Shopify's billing location report shows sales by the country or region on each order's billing address. If many regions sell steadily, split them at random into test and control. If a few regions carry everything, Google's researchers built a time-based method for that case. Path: Shopify admin > Analytics > Reports > Category filter > Sales > Total sales by billing location.
Ask Google Ads to run it for you. Conversion Lift based on geography supports Search campaigns and splits the country into Google Marketing Areas. It shows a feasibility status before you commit, but it isn't available for every account. Path: Google Ads > Goals > Lift measurement > plus button > Conversion Lift under Based on Geo.
Or switch the test regions off yourself. In the brand campaign, target the control regions only and set Location options to Presence, which Google's geo guide uses to prevent leakage. If the budget is capped, cut it to the control regions' past share of spend. Path: Google Ads > Campaigns > brand campaign > Settings > Locations, then Location options > Presence.
Freeze everything else. Google's guide says to avoid changes during a study and to keep other lift studies off the same campaigns. Keep promotions, email sends and other channels the same in both groups, and log the start date. Path: Google Ads > Campaigns > Change history, to confirm nothing moved.
Run past your slowest buyers. Read Days to key event for paths of two or more touchpoints, then run the test at least that long. When it ends, put the brand ads back and keep counting through a cooldown. Path: GA4 > Advertising > Key events dropdown > Key event attribution paths > Path length filter.
Score totals, not the campaign. Compare how Shopify sales moved in test and control regions from baseline to test. Do the same for paid plus organic brand clicks. Divide the sales gap by the brand spend you saved, and that is your incremental return. Path: Shopify admin > Analytics > Reports > Total sales by billing location, then Google Ads > Report editor > Basic tab.

A worked example

Two facts from one store's export set the frame. On its Journeys sheet, journeys of 4 to 9 touches took 16.9 days to buy. On its Break-even sheet, a 40% margin gives a break-even ROAS of 2.5x, since 1 divided by 0.40 is 2.5. The first number tells you how long to keep counting. The second is the line brand ads must clear, and it is arithmetic on a margin, not a ROAS anyone measured.

Now a test, with round numbers made up for illustration.

For illustration	Test regions	Control regions
Weekly sales, baseline weeks	€20,000	€20,000
Weekly sales, test and cooldown weeks	€20,400	€21,000
Weekly brand spend, baseline weeks	€1,000	€1,000
Weekly sales Google Ads credited to brand ads, baseline	€12,000	€12,000
Paid brand clicks a week, test weeks	0	2,000
Organic brand clicks a week, test weeks	2,700	1,000

Say the two groups each sold €20,000 a week before the test. Say the control regions then sold €21,000 a week during the test and cooldown, a seasonal lift of 5%. If the test regions had kept pace, they would have sold €21,000 too, but they sold €20,400. In this worked example, the pause cost €600 of sales a week.

Say the brand ads in the test regions used to cost €1,000 a week. In this worked example, their incremental return is €600 divided by €1,000, or 0.6x, far below the 2.5x line. If the campaign report had credited them with €12,000 a week, only 5% of what they claimed was extra.

The clicks tell the same story sooner. In this worked example, organic brand clicks in the test regions rose from 1,000 to 2,700 a week while paid clicks fell to zero. For illustration, that keeps 90% of the 3,000 combined brand clicks, which is why sales barely moved.

Flip one input to see the other side. If the gap had been €3,000 a week, the return would be 3x and the brand ads would pay their way. Before you cut, repeat the test once in other regions.

What to check when the result looks wrong

Paid brand clicks still show in the test regions. Google says 100% accuracy is not guaranteed, because location targeting runs on signals. Check that Location options say Presence, then open Campaigns > Insights & reports > When and where your ads showed > Matched locations.
Both groups lost traffic when the test began. Google's guide warns that moving from country targeting to city or regional targeting reduces traffic, since it cannot place every user precisely. Compare test with control, never with last month.
Search Console shows no regional difference. Its country dimension groups clicks by the country where the search started, so it cannot split one country into regions. Use it for the national trend and Shopify billing regions for the comparison.
The Branded filter has gaps. The filter only holds data from 11 March 2025, when Google introduced it, and it is missing for sites with few impressions. Use Queries containing your brand name instead.
Control regions spent more than usual. The capped budget leaked into them. Cut the daily budget to the control regions' share, as Google's guide sets out, and restart the clock.
Performance Max spend jumped during the test. It found your brand searches. Check that the brand exclusion is attached, then restart.

What to do this week

List every campaign that shows on your name. In Google Ads, open Campaigns, then Insights and reports, then Search terms, and filter for your brand. Pass: only the brand campaign appears. Fail: Performance Max or a generic campaign shows up too, so attach a brand exclusion before any test.
Count your usable regions. In Shopify admin, open Analytics, then Reports, filter the Category to Sales and open Total sales by billing location. Pass: several regions sell in most weeks. Fail: one region carries nearly all sales, so plan alternating weeks instead of a regional split.
Check whether Google will run the test. In Google Ads, go to Lift measurement in the Goals menu and select the plus button. Pass: Conversion Lift appears under Based on Geo and the feasibility status reads High. Fail: it is missing or reads Low, so run the manual version from step 6.

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: About the search terms report (Google Ads Help); Apply brand exclusions to Performance Max or Search campaigns (Google Ads Help); Performance report (Search results): Common tasks and use cases (Search Console Help); Performance report (Search results): Advanced filtering and comparison (Search Console Help); Performance report (Search results): Dimensions and data groupings (Search Console Help); About measuring paid and organic search results (Google Ads Help); Sales reports (Shopify Help Center); Measuring Ad Effectiveness Using Geo Experiments (Google Research); Estimating Ad Effectiveness using Geo Experiments in a Time-Based Regression Framework (Google Research); Set up Conversion Lift based on geography (Google Ads Help); Implement campaigns for geo experiments (Google Ads Help); About change history (Google Ads Help); Key events attribution paths report (Google Analytics Help); Target ads to geographic locations (Google Ads Help); View matched locations and distance reports (Google Ads Help)

Frequently asked questions

Should I exclude the test regions or target the control regions only?
Either keeps brand ads out of the test regions. Google's geo guide targets the control regions only and sets Location options to Presence to prevent leakage. Whichever you pick, check the budget, because a capped campaign can push the money it saves into the regions that still run ads.
Can I test brand Shopping ads the same way?
Yes, but score them elsewhere. Google's paid and organic report counts text ads only, not Shopping ads. Fence Shopping and Performance Max with brand exclusions, then judge the test on Shopify sales by billing region rather than on any single ad report.
What if my regions sell very unevenly?
Pair regions that moved together in the baseline weeks, and judge each group against its own history, not against the other group's size. If one region carries most sales, a time-based design built for few regions fits better than a random split.

Go deeper: Causal attribution, explained.

Sixty-second versions of these ideas: Causality Engine on YouTube Shorts.

Keep reading

Terms in this article

Browse the full glossary