The additional outcome caused by an intervention compared with an estimate of what would have happened without it.
Incrementality concerns the additional outcome caused by an intervention compared with what would have happened without it. For SEO, the question might be whether a group of content improvements produced additional qualified inquiries, beyond changes caused by demand, other marketing or the business’s normal trajectory.
The unobserved alternative makes this harder than comparing two totals. A before-and-after rise may coincide with seasonal demand or a new promotion. Controlled experiments aim to create a more credible comparison between exposed and comparison groups. Google Ads Conversion Lift illustrates that experimental principle for advertising; it is not an automatic way to test an organic SEO program.
An SEO evaluation might compare appropriately selected page groups or markets, but a credible design needs attention to pre-existing differences, spillovers, timing and other changes. Small samples and noisy outcomes may leave the result inconclusive. Attribution answers which observed touchpoints receive credit under rules or a model; incrementality asks about causal additional impact. Those numbers can differ without either report containing an arithmetic error.
Takeaway: Separate growth observed after SEO work from growth that the evidence can reasonably attribute to the work.
Example
A business records 120 inquiries after a change, versus 100 beforehand. A credible comparison suggests inquiries would have reached 115 anyway. The estimated incremental difference is five, subject to the comparison’s assumptions and uncertainty, not all 20 additional inquiries.
What to check
Define the intervention, comparison, outcome and analysis period before evaluating the result. Report uncertainty and competing explanations instead of forcing a positive causal conclusion.
For more detail, see About Conversion Lift.