Real-World Incrementality Results
What good and bad incrementality looks like in practice
In practice, incrementality results vary dramatically:
10-30% incrementality (most conversions would happen anyway)
30-50% incrementality (some lift but significant organic overlap)
50-70% incrementality (clearly driving growth beyond organic)
70%+ incrementality (strong causal impact)
A campaign with 6x incrementality (600% lift) means the conversion rate in the exposed group is 6x higher than the control group. These are the campaigns worth scaling aggressively.
The Challenge and Specify's Solution
A major challenge with traditional incrementality testing is inefficiency.
To measure lift, advertisers typically withhold ads from a control group by design. Those users still exist in the target audience, but are intentionally not shown ads.
The result is that a significant portion of a campaign — often 10% to 30% — is reserved solely for measurement rather than conversion, despite targeting users who would otherwise be eligible.
In web2, major advertising platforms (Google, Meta, etc.) offer incrementality testing primarily to established advertisers with sufficient scale, budget, and historical data. For many teams, these tests can be expensive, slow to run, and operationally complex.
Specify approaches incrementality differently. Due to the transparent nature of onchain transactions, Specify can construct statistically comparable control cohorts using the same targeting, clustering, and eligibility logic — without permanently withholding impressions from users who are eligible to be shown ads on participating publishers.
“Incrementality is a far harder question, but you should be asking it, since it is the most important question in marketing. Today, however, too many marketers rely on post-click conversions without asking, 'Is this marketing incremental or are we just tracking things that would have happened anyway?'”— Alex Schultz, Click Here(source)