Conversion Lift Study
Paid MediaAlso: Lift Test · Incrementality Test · Holdout Test
Quick definition
A conversion lift study is a controlled test that measures how many extra conversions an ad campaign actually caused, compared to what would have happened without it. It works by holding back a group of people from seeing the ad and comparing their conversion rate against the group that did see it.
How it varies across Australia
Reported lift and true incremental lift diverge more in mature Australian paid media accounts than in newer ones, because a growing share of exposed conversions would have happened anyway through direct or organic channels. The gap tends to widen as brand awareness grows.
Explore acquisition benchmarks across Australian industries →What it actually means
Attribution and lift answer different questions, and confusing them is expensive. Attribution asks who gets credit for a conversion that already happened. A conversion lift study asks whether the conversion would have happened at all without the ad.
The method borrows from clinical trials. You split your audience into two groups at random. One group sees the campaign. The other, the holdout, doesn't. If the exposed group converts at a meaningfully higher rate than the holdout, the difference is your incremental lift, the conversions the campaign actually caused rather than simply claimed.
This matters because CPA and ROAS as reported by ad platforms count conversions that would have happened anyway. Someone who already intended to buy and searched your brand name gets counted as a paid search win even though the ad added nothing. A lift study strips that noise out and shows the honest contribution of the channel.
It's not free. Running a proper holdout costs you conversions during the test, requires enough volume to reach statistical significance, and takes real coordination with your ad platforms. Most Australian businesses skip it entirely and run on attribution alone, which is fine until a channel that looks brilliant on paper turns out to be adding almost nothing.
Attribution tells you who got credit. A lift study tells you whether the sale would have happened without the ad at all.
How to calculate it
Lift = conversion rate of exposed group minus conversion rate of holdout group
Worked example. 10,000 people see the campaign and 500 convert, a 5% rate. A matched holdout of 10,000 people don't see it and 350 convert, a 3.5% rate. Lift is 1.5 percentage points, meaning roughly 30% of the exposed group's conversions were incremental and the rest would have happened anyway.
The Australian context
Smaller audience sizes across most Australian categories make it harder to reach statistical significance quickly. A lift study that would resolve in two weeks in a large United States market might need six to eight weeks locally to collect enough conversions in both groups. Businesses running lift tests here need to plan for longer test windows and budget accordingly rather than pulling the plug early because the result looks inconclusive.
Where people get this wrong
Related terms
Common questions
What is a conversion lift study used for?
It's used to prove whether a campaign is causing conversions rather than just claiming credit for conversions that would have happened anyway. Businesses run them to validate whether a channel deserves its budget or is riding on demand created elsewhere.
How long does a conversion lift study take?
It depends on conversion volume. High-traffic campaigns can resolve in one to two weeks. Lower-volume Australian campaigns often need six to eight weeks to collect enough conversions in both the exposed and holdout groups to reach statistical significance.
Can small businesses run a conversion lift study?
It's possible but harder. Lift studies need enough conversion volume in both groups to detect a real difference. Businesses with low conversion counts often find the test runs for months without reaching a reliable result.
Is a conversion lift study the same as A/B testing?
They're related but not identical. A/B testing usually compares two versions of something shown to everyone. A lift study compares an exposed group against a holdout group that sees nothing from that campaign at all, isolating the causal effect of running the campaign in the first place.
Debrief
Get the next one
No spam. No fluff. Just the next article, straight to your inbox.
Keep exploring
About New Rebellion
New Rebellion is a marketing intelligence consultancy. We build tools, score Australian businesses on how their marketing actually performs, and publish Debrief every day. This dictionary is part of how we work in the open.
How we think →