Geo-Lift Test
A causal experiment that runs a campaign in some geographies and not others, then measures the difference.
Last updated
Definition
Geo-lift is the gold-standard incrementality measurement when user-level holdouts are infeasible. Pick treatment + control geos with similar pre-period trends, run the campaign in treatment only, measure the lift via Synthetic Control or DiD. Resistant to platform-attribution bias because it doesn't depend on platform-reported conversions.
How it applies in India
India is well-suited for geo-lift because the metro-vs-tier-2 split provides natural cohorts; works well for performance + brand channels both.
An example
A brand that delivers to twelve towns wants to know what its Meta and Google ads really add. For two weeks it switches the ads off in one mid-sized town and keeps them running everywhere else. If orders in that town hold up while the rest carry on as before, the ads were adding less than the ad platforms claimed; if orders there fall away, the gap shows what the ads were doing.
How WatEase uses it
WatEase runs this as the quarterly holdout. It takes about one town in ten (at least one and never all, never the home town, and not the busiest one when there are three or more) out of every live Meta and Google campaign for 14 days, then puts them back. The first test runs 28 days after a business’s first campaign goes live and after that at most once a quarter; it skips one-town businesses, campaigns still in their learning days and accounts in watch-only mode. The owner is told seven days ahead, or asked to approve, and can switch it off. Once both sides have enough orders, the result reads “Ads added about N% more orders, ±M%”, worked out as a difference in differences.
Frequently asked questions
Will WatEase pause my ads in some towns without telling me?
No. It tells you seven days before a test, or asks for approval if you prefer, and you can switch the quarterly holdout off. The pause lasts 14 days and covers about one town in ten.
Related terms
- IncrementalityThe lift caused BY a marketing intervention, vs. what would have happened without it.
- Synthetic Control Method (SCM)Builds a counterfactual for a treated unit from a weighted combination of untreated units.
- Causal InferenceStatistical methods that estimate causal effects from observational or experimental data.
Try WatEase
Run your WhatsApp commerce on the platform built for India — Cloud API, GST invoices, UPI checkout, opt-in tracking, and 0% markup on WhatsApp messages (competitors add 20–25%). Pricing is on request — quoted by the WatEase team.
Get Started Today