Glossary · Measurement
Incrementality
The share of conversions that happened because of the spend, against those that would have occurred anyway. High reported ROAS with low incrementality means you are paying for sales you already had.
What Incrementality means
Incrementality is the share of results that happened because of a piece of marketing, as opposed to the results that would have happened anyway. A holdout test makes the split visible: if a campaign is credited with 1,000 sales and 850 of those buyers would have bought without it, the campaign's incremental contribution is 150 sales, and those 150 are the only ones its budget actually bought.
It answers a different question from attribution. Attribution asks which touchpoint a sale should be credited to; incrementality asks whether the sale needed the touchpoint at all. Every platform can attribute a sale to itself, and none of them can tell you, from inside its own dashboard, whether the customer would have come anyway. Only a comparison against a group that did not see the spend can.
The Incrementality formula and a worked example
The formula and an illustration; the corpus line is the audited figure behind it.
Incremental share = (exposed conversions − holdout conversions) ÷ exposed conversions
Worked example Illustration, not a client figure
A geo holdout: exposed regions convert 1,000 times, matched holdout regions 850. Only 150 of the 1,000 conversions, 15%, were incremental. The platform reported all 1,000.
Why Incrementality matters in an audit
Incrementality is the test behind most of the audit's paid media findings. The method is a holdout: pause the spend in matched regions, or withhold it from a random share of the audience, and compare the results with and without it. The difference is the incremental result, and the share of the reported result that survives without the spend is the part the budget never needed to buy.
It matters because the campaigns with the best reported returns are often the least incremental. Branded search, retargeting, affiliate and win back campaigns reach people who are already on their way, so they convert well and add little. Non incremental paid spend appeared in 68% of the 74 audits, the most common finding in the corpus, and every one of those findings rests on an incrementality test.
The idea reaches beyond media. A loyalty scheme is incremental only if the repeat purchases it rewards would not have happened anyway; an annual plan discount is worth its cost only if it prevents more churn than it gives away; a sponsorship pays only if it moves something measurable. The published cases below apply the same test to each.
Incrementality in the audited cases
Published cases where a holdout or a lift test decided the finding, in media and beyond it, each with the audited annual figure.
- Retail
Online ads were paying for in store demand
−$88k a year, findings in 7 days.
Read the case → - DTC / Ecommerce
The agency’s ‘winning’ campaign was the biggest loser
−$53k a year, findings in 7 days.
Read the case → - Retail
Loyalty points cost more than the repeat they drove
−$71k a year, findings in 7 days.
Read the case → - Subscription
Annual plans were discounted below their retention value
+$57k a year, findings in 6 days.
Read the case → - Pharma
Detailing and sample spend produced no prescribing lift
−$64k a year, findings in 7 days.
Read the case → - Pharma
Congress and KOL sponsorship showed no measurable return
−$88k a year, findings in 7 days.
Read the case → - Subscription
Winback campaigns were rebuying customers who’d have returned free
−$33k a year, findings in 5 days.
Read the case → - DTC / Ecommerce
Affiliate commissions were being paid on sales the brand already owned
−$44k a year, findings in 5 days.
Read the case →
Questions about Incrementality
What is incrementality in marketing?
Incrementality is the share of conversions that happened because of the marketing, measured against a comparison group that did not see it. Conversions that would have happened anyway are not incremental, however a platform attributes them.
How do you measure incrementality?
With a holdout. Withhold the spend from matched regions or a random share of the audience and compare the results. If exposed regions convert 1,000 times and matched holdout regions 850, 150 conversions were incremental, an incremental share of 15%, while the platform reported all 1,000.
What is the difference between incrementality and attribution?
Attribution decides which touchpoint gets the credit for a sale. Incrementality asks whether the sale needed the touchpoint at all. A campaign can be credited with every sale it touched and still be mostly non incremental, which is why the audit tests rather than reattributes. Read the non incremental spend page →
Know the term. Now measure it in your own numbers.
Incrementality, measured against your own accounts in 5 to 7 working days. $10k to $50k of findings, or your money back.
Fixed fee. No retainer. NDA first.
Go deeper
Incrementality and the wider audit evidence
- All 14 termsThe full glossary: every term with a formula and a worked example.→
- All 74 findingsRanked by annual value, $58k median, each one a real case.→
- The benchmarkWhere 74 SMB marketing budgets leaked, by share of spend and by industry.→
- Margin leaksThe five leak patterns in depth: what each one is and how an audit finds it.→
- Marketing audit checklistThirty checks across paid media, attribution, pricing, retention and reporting.→
- What an audit costsThe 2026 market ranges with sources, against the fixed fee tiers.→