TL;DR
Incrementality is the measure of how many conversions, sales, or pipeline outcomes were directly caused by marketing activity — versus how many would have happened without it. Median attribution inflation across B2B SaaS marketing is 2–4×: attributed revenue overstates true incremental revenue by that multiple. Measuring incrementality is how you find the overstatement.
What is incrementality?
Incrementality (also called incremental lift, causal impact, or true ad effectiveness) measures the additional conversions, revenue, or pipeline attributable to a specific marketing action — above and beyond what would have occurred naturally. It answers: "did this campaign cause these conversions, or would they have happened anyway?"
Incrementality differs from attribution. Attribution models (last-touch, first-touch, linear, U-shaped) divide credit among observed touchpoints based on rules or algorithms. None of them have a counterfactual. Incrementality measurement uses controlled experiments — holdout tests or geo-lift tests — to establish what conversion rate looks like without the campaign, which is the only rigorous way to separate correlation from causation.
For B2B SaaS operators and D2C brands, incrementality measurement matters most for the three channels with the highest organic conversion rates: retargeting, branded search, and nurture email. These channels show high attributed ROAS precisely because the people they target were already likely to convert. Incrementality testing reveals how much of that ROAS is genuine.
Why incrementality matters for operators
Attribution inflation is ubiquitous and material. Across B2B SaaS and D2C, industry analysis consistently shows that last-click attribution overstates the true contribution of retargeting by 2–5× and branded search by 3–10×. On a $500K annual paid budget, the cumulative allocation error from un-corrected attribution can represent $100K–$200K misallocated per year.
The specific cost is compounding. Operators who scale retargeting based on attributed ROAS aren't just over-spending on one channel — they're under-spending on the prospecting channels that actually generate new demand. The attributed retargeting ROAS looks strong because it's capturing organic demand; the prospecting ROAS looks weak because it's correctly measured against last-click. Net effect: new customer acquisition slows while the existing customer base is over-advertised.
Measuring incrementality typically produces two immediate findings: the channels with the highest attributed ROAS have the most attribution inflation, and the channels with the lowest attributed ROAS (often top-of-funnel prospecting and content) have the highest true incrementality.
How incrementality is measured
- Holdout test: Withhold the campaign from 10–20% of users at random. Compare conversion rates. Best for single-platform digital campaigns where user-level randomisation is possible.
- Geo-lift test: Run campaigns in some geographic markets, withhold in matched control markets. Best for cross-channel, TV, podcast, and offline campaigns that can't randomise at the user level.
- Marketing mix modeling (MMM): Statistical regression model that estimates the contribution of each channel to revenue over time, controlling for seasonality, price, and market conditions. Best for long-term strategic allocation; not a substitute for experiment-based incrementality.
- Synthetic control method: A statistical technique that constructs a counterfactual from a weighted combination of control units. Used when geographic markets are unavailable or when the test market is too large to replicate.
Incrementality benchmarks by channel
| Channel | Typical incrementality rate | Typical attributed ROAS | True incremental ROAS | Implication |
|---|---|---|---|---|
| D2C retargeting (existing customers) | 20–40% | 4–8× | 0.8–2× | Often not worth the spend at scale |
| D2C prospecting (new audiences) | 65–85% | 1.5–3× | 1.2–2.5× | Genuinely incremental; scale with confidence |
| Branded search (existing demand) | 10–25% | 6–15× | 0.6–2.5× | Mostly organic; reduce spend, test slowly |
| Non-branded search | 50–75% | 2–5× | 1.5–4× | Strong incrementality; usually underinvested |
| B2B SaaS nurture email | 15–35% | Attributed: high | Low–moderate | Accelerates timing; doesn't create demand |
| B2B SaaS brand awareness (LinkedIn) | 25–45% | Hard to attribute | 0.5–2× | Assists pipeline; hard to measure short-term |
Sources: Meta Conversion Lift Studies 2024; Google Brand Lift Studies 2024; Common Thread Collective D2C Benchmarks 2025; Fairview customer data. Ranges are directional — actual incrementality varies by audience, creative quality, and competitive context.
Common mistakes with incrementality measurement
1. Treating attribution data as a proxy for incrementality. They measure fundamentally different things. Attribution divides credit among observed touchpoints; incrementality measures causal impact against a counterfactual. A channel can have excellent attribution and near-zero incrementality — this is the retargeting trap.
2. Measuring incrementality once and treating it as permanent. Incrementality changes as organic conversion rates shift, competitive advertising intensity changes, and audience saturation increases. Run incrementality tests at least quarterly on your 3–5 largest channels.
3. Only testing low-performing channels. The channels most likely to have significant attribution inflation are your best-performing ones — specifically retargeting and branded search. High attributed ROAS is a signal of potential attribution inflation, not evidence that incrementality is high.
4. Not calculating incremental CPA alongside lift percentage. A 30% lift sounds strong. Whether it's economically viable depends on: what is the conversion rate of the control group, and what is the ad spend per incremental conversion? Always translate lift to incremental CPA before making budget decisions.
5. Using incrementality results to cut spend without modelling the revenue impact. Cutting retargeting spend by 50% because holdout tests show 20% incrementality will reduce revenue somewhat. The right move is gradual budget reallocation — shift 10–15% at a time and monitor total revenue, not just channel revenue.
How Fairview uses incrementality data
Fairview's Margin Intelligence module accepts incrementality-adjusted ROAS inputs from holdout and geo-lift test results. When operators input a channel's measured incrementality rate, Fairview recalculates that channel's effective contribution margin — removing the attribution inflation from the analysis.
The Next-Best Action Engine surfaces allocation recommendations based on incrementality-corrected data: "Retargeting channel ROAS is 4.2× attributed but 0.9× incremental (from Q1 holdout test). Recommend reallocating $18,000/month of retargeting budget to prospecting campaigns where measured incrementality is 71%."
Companies using Fairview that incorporate incrementality measurement into their channel analysis typically identify 2–3 channels with material attribution inflation in the first quarter, enabling budget reallocation that maintains or grows total pipeline at lower spend.
→ See how Margin Intelligence tracks true channel performance
At a glance
- Category
- Marketing Metrics
- Related
- 5 terms
Frequently asked questions
What is incrementality in simple terms?
How many of your conversions actually happened because of your ad, versus how many would have happened anyway. If 100 people bought after seeing your retargeting ad, but 80 of them were already planning to buy, the incremental value of that ad is 20 conversions — not 100.
How do you measure incrementality?
With a controlled experiment — either a holdout test (withhold the campaign from a random user subset) or a geo-lift test (withhold from matched geographic markets). Both give you a control group that shows what conversion rate looks like without the campaign. The difference between exposed and control is the incremental lift.
What is a good incrementality rate?
Above 60% is strong — more than half of your attributed conversions are genuinely caused by the campaign. 30–60% is acceptable for most channels. Below 20% means most attributed conversions are organic — the channel is claiming credit, not creating it. Retargeting and branded search frequently fall below 20%.
Is incrementality the same as ROAS?
No. ROAS (return on ad spend) uses attributed revenue — it counts all conversions among people who saw an ad. Incrementality-adjusted ROAS uses only the incremental conversions from a holdout experiment. For most retargeting campaigns, attributed ROAS is 3–8× while true incremental ROAS is 0.5–2×.
How often should you run incrementality tests?
Quarterly for your 3–5 largest channels. Monthly if you're making significant budget changes. Incrementality shifts over time as organic demand changes, creative fatigues, and audiences saturate — a test that was valid 6 months ago may not reflect current reality.
Sources
- OpenView SaaS Benchmarks 2025
- Common Thread Collective D2C Benchmarks 2025
- ProfitWell Research
- Pavilion Operator Survey 2024
- Fairview customer data (B2B SaaS + D2C, 2025)
Fairview is an operating intelligence platform that connects incrementality test results to channel spend — replacing attributed ROAS with measurements that actually predict what happens when you change your budget. Start your free trial →
Siddharth Gangal is the founder of Fairview. He built the incrementality module after watching operators keep scaling their top attributed-ROAS channels into diminishing returns because nobody had run a holdout test to check whether those channels were creating demand or just taking credit for it.
See it in Fairview
Track Incrementality automatically.
14-day free trial. No credit card. First data source connected in 5 minutes.