Measurement
Incremental Revenue
Incremental revenue is the revenue a change added: what a business actually earned after the change, minus what it would have earned without it. The second figure, the counterfactual baseline, cannot be observed, so it is estimated from a control group or a forecast, and incremental revenue is stated as a range.
- Industry term
By QueryRev ResearchPublished Updated5 min read
Why it matters.
It is the only revenue figure that says what the work did. Attributed revenue says which sales a rule gives to search; incremental revenue says which sales would not have happened without a change. Advertisers measure the same thing with lift tests: Google Ads describes Conversion Lift as a way to measure the conversions directly driven by ads, by comparing people who see them with a control group who do not[1].
It is a different claim from the other kinds of revenue an SEO report carries, and it is never added to them. The guide to SEO reporting in revenue sets out the four kinds and how each is labelled.
Search rarely allows a clean experiment. The CausalImpact method was built for exactly that case: it predicts the response that would have occurred had no intervention taken place, in a synthetic control, where a randomised experiment is unavailable[2]. Its documentation builds that prediction from control series, such as clicks on sites the change did not reach[3].
Changes to who can read a site are increments too. In QueryRev’s study, AI training crawlers are blocked by name on 7.9% of the sites read on average, and AI search crawlers on 4.8%[4]. A site that opens its pages to a search crawler it used to block has made a change whose effect on AI referral traffic and revenue can be measured this way.
How it is measured.
Take the revenue recorded over a window after the change, from orders or closed deals, for the pages the change touched. Then estimate the baseline for the same window: from similar pages left unchanged, from the same pages’ history with its seasons, or from both.
Read the actual in the record of sales rather than in analytics, over a window as long as the baseline’s, so that refunds and cancelled orders do not count as a lift. Incremental revenue is the actual minus the baseline. Because the baseline is an estimate, it has a range, and so does the result: the actual minus the high end of the baseline, to the actual minus the low end.
The estimate holds only if its assumptions do. The CausalImpact documentation names two: the control series were not themselves affected by the change, and their relationship with the measured series stays as it was before the change[3].
How a baseline enters the return on search work is set out in SEO ROI: how to measure it.
Measured by
- Actual revenue
- Baseline range
- Lift as a range
A worked example.
Sample data: an invented site, to show the method.
A product page of a sample shop is rewritten after week six. Over the 28 days of the measurement window it records $62,400 of known revenue from orders. A baseline modeled from the page’s own history and from similar pages that were not changed puts what it would have earned at $55,100 to $55,900.
Incremental revenue is the $62,400 minus the baseline: a lift of $6,500 to $7,300 over the 28 days. It is reported as that range and labelled incremental, beside the $62,400 labelled known, and the two are never added together.
What to do, step by step.
Name the change, the pages it touched and the date, before the result is known.
Choose the baseline method in advance: unchanged pages of the same kind, the pages’ own history with its seasons, or both.
Measure the actual revenue from the record of sales over a fixed window after the change.
Subtract the baseline’s high and low ends to get the lift as a range.
Label it incremental, and keep it apart from known, influenced and modeled figures.
Check the assumptions: the comparison pages were not touched by the change, and nothing else changed in the same weeks.
Mistakes to avoid.
Calling attributed revenue incremental. Sales a rule gives to search would partly have happened anyway. Incremental revenue needs a baseline.
Writing the lift as one number. The baseline is an estimate, so the lift is a range. A midpoint alone claims a precision nobody has.
Picking the baseline after the result. A baseline chosen once the numbers are in can be made to show almost any lift. Fix the method first.
Measuring too early. A window that closes before search engines have read the changed page again measures the delay, not the change. Give it the weeks it needs.
Using comparison pages the change reached. If the comparison pages were affected, through internal links or a shared template, the estimate can be wrong in either direction[3].
How QueryRev treats it.
QueryRev measures incremental revenue only for changes that went through verification: the change is written to the Change Ledger with its date, a baseline range is set before the window opens, and the lift is reported as a range labelled incremental. The methodology page works the same sample through.
Questions people ask.
What is the difference between incremental revenue and attributed revenue?
Attributed revenue is the share of sales that a rule gives to a channel. Incremental revenue is the extra revenue a change caused, measured against a baseline of what would have happened without it.
Can incremental revenue be negative?
Yes. If actual revenue falls below the baseline, the change cost revenue, and the range says how much. A range that runs from negative to positive means the effect cannot be told apart from zero.
How long should the measurement window be?
Long enough for the change to take effect and for ordinary swings from week to week to even out, and fixed before the result is known. For most page changes that means weeks, not days.
Sources and citation.
Primary sources only: the search engines, the AI vendors, standards bodies and QueryRev’s own research. Each is cited in the text by its number.
About Conversion LiftGoogle Ads Help ·
CausalImpact: an R package for causal inference using Bayesian structural time-series modelsGoogle ·
State of AI crawler access, October 2026QueryRev Research ·
Cite this definition
A suggested citation. Its link goes to the definition, whose anchor does not change.
QueryRev Research. “Incremental Revenue.” QueryRev AI glossary, 19 October 2026. https://queryrev.com/glossary/incremental-revenue#definition
Related terms