Pausing brand ads lost almost nothing
A Kozhikode restaurant chain spent ₹40,000 a month on Google Ads for its own brand name, with a reported ROAS of 12.
They paused brand ads in two of four cities for a month. Orders in those cities fell by less than 3%, because people clicked the organic listing instead.
They moved most of that budget to non-brand and Meta prospecting, where the incremental gain was far larger.
Illustrative example based on a common pattern in Kerala e-commerce accounts. Numbers rounded.
What incrementality is
Attribution asks “which ad gets credit?”. Incrementality asks “would this sale have happened anyway?”. The difference matters most for retargeting, brand search and loyal customers.
Meta and Google offer conversion lift studies for eligible accounts. Smaller businesses can run geo holdouts: turn ads off in some comparable areas and compare.
Incremental lift = (test group result − holdout group result) ÷ holdout group result
Ways to measure incrementality
Methods
| Method | Good for |
|---|---|
| Platform conversion lift study | Larger Meta or Google accounts that meet eligibility |
| Geo holdout (pause ads in some districts) | Local and regional businesses |
| On/off test over time | Very small accounts, with care for seasonality |
| Marketing mix modelling (MMM) | Large multi-channel budgets |
How experienced marketers use Incrementality (lift test)
1. Test brand search spend
Find out whether brand ads add sales or just take credit.
2. Right-size retargeting
Retargeting often reports high ROAS but adds less than it claims.
3. Justify prospecting
Show owners that top-of-funnel ads grow total sales.
What good looks like
| Business type | Starting rule | Why |
|---|---|---|
| Test length | 2–4 weeks minimum | Longer for long sales cycles. |
| Holdout size | 10–50% of audience or areas | Bigger holdouts give clearer results. |
| Comparable areas | Similar size and history | Match districts before testing. |
When the number lies
- Too small to tell
- Small tests produce noise. If the difference is within normal week-to-week change, it’s inconclusive.
- Outside events
- Festivals, competitors and weather can swing a holdout test.
Mistakes to avoid
Trusting platform ROAS as proof
Budgets go to ads that take credit, not ads that cause sales.
Stopping tests early
Random noise is mistaken for a result.
Testing during a festival
Results reflect the season, not the ads.
Related terms
Questions people ask about incrementality
What is incrementality in marketing?
The extra results caused by marketing that wouldn’t have happened without it.
What is a lift test?
An experiment comparing a group exposed to ads with a holdout group that isn’t, to measure the difference ads made.
Can small businesses test incrementality?
Yes, with simple geo holdouts or carefully timed on/off tests, though results are less precise.
Is retargeting incremental?
Partly. Many retargeted buyers would have bought anyway, so test before scaling it.
Put it into practice
- SOP 307: Create a Monthly Google Ads Report Clients Actually Read →
- SOP 124: Audit a Meta Ads Account →
Prefer to talk first? Book a free counselling call.