What Is Incremental Lift? A Plain-English Guide
. By Ads Daddy.
# What Is Incremental Lift? A Plain-English Guide  **Incremental lift is the extra conversions or revenue your ads caused, over and above what would have happened anyway.** It's the difference between "people who bought after seeing my ad" and "people who bought *because* they saw my ad". Those are not the same crowd. Think of it like dating. Someone who was already going to marry you doesn't need a rose bouquet. Lift tells you whether the bouquet worked, or whether you just paid for flowers on the way to a yes you already had. ## TL;DR - Incremental lift = the extra results caused by your ads, measured against a control group that didn't see them. - Relative lift is a percentage. Absolute lift is the number of extra conversions. Watch both, because the second one pays the bills. - Real lift is often modest, sometimes single digits for conversions. That's normal, not a failure. - Lift doesn't replace attribution or marketing mix models. It calibrates them. - You don't need a giant budget to test any more. Start with one campaign, one clear question. ## Table of Contents - [So what does incremental lift actually mean?](#so-what-does-incremental-lift-actually-mean) - [The incremental lift formula](#the-incremental-lift-formula) - [Relative lift vs absolute lift: don't get fooled](#relative-lift-vs-absolute-lift-dont-get-fooled) - [What's a "good" lift number?](#whats-a-good-lift-number) - [Certainty of lift: is it real or just luck?](#certainty-of-lift-is-it-real-or-just-luck) - [How to measure incremental lift: the main methods](#how-to-measure-incremental-lift-the-main-methods) - [How to run your first test: 7 steps](#how-to-run-your-first-test-7-steps) - [Lift vs attribution vs marketing mix models](#lift-vs-attribution-vs-marketing-mix-models) - [Where lift testing fits for small business](#where-lift-testing-fits-for-small-business) - [Common mistakes (and how to dodge them)](#common-mistakes-and-how-to-dodge-them) - [Our perspective: lift is the honest friend in your reporting](#our-perspective-lift-is-the-honest-friend-in-your-reporting) - [Want someone else to run the numbers?](#want-someone-else-to-run-the-numbers) - [FAQ](#faq) ## So what does incremental lift actually mean? Every ad platform loves to take credit. Someone searches your brand name, clicks an ad, buys. The dashboard cheers. But that person was already looking for you. Incremental lift strips out the freebies. It asks one blunt question: **what would have happened if we'd done nothing?** To answer it, you split an audience in two: 1. **Treatment group:** sees your ads. 2. **Control group:** doesn't (or sees a placebo, like a public service ad). Both groups are picked at random, so they're alike in every way except the ads. Any gap in conversions between them is the lift. The gap is your ads' real fingerprint. You'll also hear it called *incrementality*. Same idea. Incrementality is the concept. Lift is the number you get out of it. If you want the wider concept first, we've covered it in [what ad incrementality means for marketers](https://adsdaddy.com/au/blog/what-is-ad-incrementality). ## The incremental lift formula Two formulas matter. Learn both and you'll out-think most dashboards. **Relative lift (%)** = (Treatment conversion rate − Control conversion rate) ÷ Control conversion rate **Absolute lift (extra conversions)** = (Treatment conversion rate − Control conversion rate) × Number of people in the treatment group Relative lift tells you how much better the ad group did in proportion. Absolute lift tells you how many extra customers you actually got. ### A worked example Say you run a Shopify store selling pet gear. You test a Meta campaign with 100,000 people in the treatment group and 100,000 in the control group. | | Treatment (saw ads) | Control (no ads) | |---|---|---| | People | 100,000 | 100,000 | | Purchases | 1,200 | 1,000 | | Conversion rate | 1.2% | 1.0% | - **Relative lift:** (1.2% − 1.0%) ÷ 1.0% = **20%** - **Absolute lift:** 0.2% × 100,000 = **200 extra purchases** Now here's the ROI bit. The platform's attribution might have claimed all 1,200 purchases. Only 200 were truly caused by the ads. The other 1,000 were coming anyway. If you spent A$10,000 and each sale is worth A$100 in revenue, the ads *caused* A$20,000 in revenue, not A$120,000. That's still a 2x return on revenue, before costs. Good, but a very different story from the dashboard's version. (Numbers above are made up to show the maths. Your results will differ.)  ## Relative lift vs absolute lift: don't get fooled Headlines love relative lift. "Ads boosted conversions 300%!" Sounds wild. But 300% of almost nothing is still almost nothing. Go from 2 sales to 8 and you've got 300% lift and six extra sales. Meanwhile a 5% lift on a huge audience might mean hundreds of extra customers. **Rule of thumb:** relative lift gets attention. Absolute lift drives ROI. Make your team report both, every time. ## What's a "good" lift number? Here's the candid bit. Chasing big double-digit lift is a classic mistake. Ads often have subtle effects. The best-known academic evidence is a Wharton study of 432 field experiments on display advertising. It found campaigns increased site visits and conversions, with median lifts of 17% for site visits and 8% for conversions. That's a median, so plenty of campaigns sat lower. The same research points to something every owner should tattoo on their forearm: **the funnel leaks.** A 10% lift in visitors often becomes only a 5 to 7% lift in buyers. If you assume it's 1:1, you'll overestimate your results. Also, effects vary a lot between businesses. What works for one e-commerce brand can be flat for another. So there's no universal "good" number. Your benchmark is *your own* previous tests. A low first result isn't a verdict that ads don't work. It's the first honest number you've had.  ## Certainty of lift: is it real or just luck? A lift number without confidence is a horoscope. Google reports something it calls certainty of lift. It's essentially 1 minus the p-value, a statistical measure of how likely your result is to be real rather than random noise. As a guide: - **90% or higher:** very good chance your ads drove the lift. - **Below 50%:** reported as "no lift", because the result isn't reliable. - **In between:** directional. Treat it as a hint, then retest or run longer. Small audiences and short tests produce noisy results. If your lift looks amazing but certainty is 60%, don't spend the money yet. ## How to measure incremental lift: the main methods You've got a few ways in. Pick based on your channel and your budget. ### 1. Randomised audience tests (conversion lift studies) The platform randomly holds back part of your audience and compares conversions. Meta and Google both offer versions of this. It's the cleanest method when you can use it. ### 2. Geo experiments No user-level splitting. Instead, you run ads in some regions and pause them in comparable ones, then compare results. Handy when privacy limits or offline sales make user-level tracking messy. For an Australian business, think matched states or metro areas. ### 3. Holdout tests you run yourself Simple version: pause a campaign for a set audience or period and see what happens to sales. It's rougher because seasonality and promos muddy the waters, but it's better than flying blind. ### 4. Ghost or placebo ads The control group sees an unrelated ad, so you compare people who were genuinely eligible for exposure. More work, tighter data. ## How to run your first test: 7 steps 1. **Pick one question.** "Do my retargeting ads cause extra purchases?" beats "Do ads work?" 2. **Choose one primary metric.** Purchases, qualified leads, bookings. Not clicks. 3. **Randomise properly.** No cherry-picking the control group. Random or it's worthless. 4. **Set a duration and stick to it.** Don't peek at day three and call it. 5. **Keep everything else steady.** No surprise sale, new landing page and budget doubling mid-test. 6. **Check certainty.** Aim for 90% or higher before acting on the number. 7. **Decide in advance what you'll do.** If lift is under X, cut spend. If over Y, scale. Write it down first, before ego gets involved. Budget used to be the big barrier. Google Ads lowered the minimum spend for its incrementality testing to $5,000 as of November 2025, down from upwards of $100,000, so it's within reach for smaller advertisers. Check the current requirements in your account before you plan, as thresholds can change. ## Lift vs attribution vs marketing mix models Here's where people get tangled. Lift doesn't replace anything. It's a calibration tool. | Method | What it tells you | Weak spot | |---|---|---| | Platform attribution | Which touchpoints got credit | Takes credit for sales that were coming anyway | | Marketing mix model | How channels and external factors (seasonality, economy) shape sales overall | Correlation-based, slow, needs lots of data | | Incremental lift test | What the ads actually caused | One channel, one moment in time | The smart move is a hybrid. Use lift results to feed real, causal data back into your attribution and your models. Then your day-to-day reporting stops being fiction. If attribution is the part that confuses you, start with our guide to [attribution modelling and ROI](https://adsdaddy.com/au/blog/what-is-ad-attribution-a-2026-guide-for-marketers). ## Where lift testing fits for small business Be honest with yourself. If you're spending A$1,500 a month, you probably don't need a formal lift study yet. You need basics: - Conversion tracking that actually works. - A sensible budget. Our walkthrough on [setting a Google Ads daily budget](https://adsdaddy.com/au/blog/google-ads-budget) shows how to work it backwards from your goals. - A clean read on [your conversion rate](https://adsdaddy.com/au/blog/understand-conversion-rate-boost-business-growth-and-ad-roi). - A habit of [testing ads](https://adsdaddy.com/au/blog/split-testing-in-ads-a-guide-for-smb-marketers) against each other. Once those are humming and you're spending real money across channels, lift testing is how you find out which campaigns are earning their keep. ## Common mistakes (and how to dodge them) - **Testing too small.** Not enough people means noisy results. Certainty tanks. - **Ending early.** Day-three excitement is not a result. - **Changing things mid-test.** New offer, new creative, new budget? You've broken the experiment. - **Judging brand search by last click.** Brand terms are the classic freebie. Test them. - **Treating one test as gospel.** Ad effects vary by brand, season and audience. Repeat the test. - **Ignoring absolute numbers.** A big percentage on a tiny base is a party trick. ## Our perspective: lift is the honest friend in your reporting We manage campaigns across Meta, Google, YouTube, LinkedIn and Microsoft Ads, and here's what we keep seeing. Owners get dazzled by platform numbers that look glorious and then wonder why the bank balance doesn't agree. Lift is the friend who tells you the truth over coffee. It's not always flattering. Sometimes it says your retargeting is mostly claiming sales that were already in the bag. That's a gift. You move that money somewhere that actually grows the business. Our view is simple. You shouldn't have to become a statistician to run ads. Our job is hands-on management, with tracking and lead delivery wired together, so the numbers you see line up with the customers you get. Tests are a tool, not a religion. Run them when the spend justifies it, and judge every campaign by extra customers and extra profit, not by applause from a dashboard. > *— Adrian* ## Want someone else to run the numbers? If you'd rather spend your week running the business than arguing with ad platforms, that's what we do. We build, manage and optimise campaigns on [Meta, Google, YouTube, LinkedIn and Microsoft Ads](https://adsdaddy.com/au), and we set up tracking so you know what's actually working. Need enquiries rather than clicks? Have a look at our [lead generation service](https://adsdaddy.com/au/lead-generation). For finance brokers, our [pay-per-lead options](https://adsdaddy.com/au/pay-per-lead) start from 70 A$ one-off per lead for car finance and from 170 A$ one-off per lead for ABN finance. **Ready to find out which of your ads really earn their keep? Book a call and let's look at your numbers together.** ## FAQ ### What is incremental lift in simple terms? It's the extra sales, leads or sign-ups that happened because of your ads, compared with a similar group who didn't see them. It separates real impact from people who would have bought anyway. ### How do you calculate incremental lift? Subtract the control group's conversion rate from the treatment group's, then divide by the control rate. That gives relative lift as a percentage. Multiply the rate difference by the treatment group's size to get absolute lift, the number of extra conversions. ### What is a good incremental lift percentage? There's no universal figure. A well-known Wharton analysis of 432 experiments found median lifts of 17% for site visits and 8% for conversions, and conversion lift is often smaller than people expect. Compare against your own past tests and check the confidence level before trusting any number. ### Is incrementality the same as attribution? No. Attribution assigns credit to touchpoints along the customer's path. Incrementality tests whether the ads caused the result at all. Use lift tests to sanity-check and calibrate your attribution. ### Can a small business run a lift test? Yes, if you have enough volume for a reliable result. Google lowered its minimum spend for incrementality testing to $5,000 in late 2025. If your budget is smaller, start with tracking, budgeting and simple holdout tests first. ## Recommended - [What is ad incrementality: the marketer's real guide](https://adsdaddy.com/au/blog/what-is-ad-incrementality) - [Attribution Modelling: Boosting ROI for Marketers](https://adsdaddy.com/blog/attribution-modelling-marketing-roi) - [Ad Quality Score Explained: What It Is and How to Lift It](https://adsdaddy.com/au/blog/explain-ad-quality-score) - [Brand Lift: How It Drives E-Commerce Growth](https://adsdaddy.com/au/blog/brand-lift-ecommerce-growth-au)