Incrementality
Affiliate programme incrementality
Assess whether affiliate activity adds orders, customers or contribution beyond what would have happened anyway, using suitable comparisons and clear limits.
Affiliate programme incrementality is the additional business outcome caused by affiliate activity. Credited sales answer a different question: which orders received affiliate credit under the programme’s rules.
To decide whether the programme merits investment, define the outcome, compare it with a credible alternative without the activity, and account for the cost difference.
Define the decision and outcome
Choose an outcome that could change a decision: additional approved orders, first-time customers, sales of a selected range or contribution. Specify the customers, products, publishers, dates, currency and treatment of discounts and refunds. Apply the same definition to the comparison group.
| Measure | What it shows |
|---|---|
| Attributed orders | Orders credited under the configured affiliate rule |
| Customer and journey indicators | The recorded mix and sequence of interactions |
| Estimated incremental outcome | The difference attributed to the activity by a suitable comparison, with uncertainty |
A first-time buyer might have bought without a publisher. An early recorded publisher touch might have helped, but the path does not reveal what the customer would otherwise have done. Use these indicators to choose questions to test, not as estimates of additional sales.
Attributed Orders vs. Estimated Incremental Outcome
- Attributed OrdersOrders credited under the affiliate programme’s rules
- Estimated Incremental OutcomeDifference in outcomes between groups with and without affiliate activity, accounting for uncertainty
Match the comparison to the investment decision
A partnership-level holdout can provide evidence about a tested activity. See Awin’s guide to testing for incrementality for setup and evaluation details.
A programme-wide claim needs evidence covering the programme or a justified combination of comparable tests. Do not apply one publisher’s result to every partner.
Where a holdout is impractical, a geographic or staged comparison may offer directional evidence if the groups and concurrent marketing are examined. A simple before-and-after total is vulnerable to promotions, stock changes and seasonality.
Steps to Assess Affiliate Programme Incrementality
- Define the outcome (e.g., first-time customers, additional contribution)
- Select a credible comparison group (e.g., holdout, geographic split)
- Measure cost difference (commissions, placements, discounts)
- Compare incremental value against incremental cost
- Use results to inform investment decisions
Compare additional value with additional cost
Report the outcome difference over the same observation period and its uncertainty. For a financial decision, compare contribution before affiliate-specific costs across groups, then subtract the difference in commission, placement payments, funded discounts and relevant programme charges.
State which costs were included so a discount or refund is not counted twice. Keep contribution from credited orders as a separate measure.
If the estimate is too imprecise to distinguish a worthwhile gain from no gain, narrow the conclusion or improve the test. Do not apply a guessed incrementality percentage to all attributed sales.
Plan investment around commercial constraints
Forecast from historic programme results, review the current commission structure and set tenancy budgets for a month or quarter ahead.
Investment terms can be adjusted to reflect commercial differences: partner numbers and types can be controlled, and commission rates can vary by profit margin or product category. When weighing those options, retain the same outcome definition and cost treatment used in the evaluation. A higher attributed-sales total alone does not show that a more generous rate or broader partner mix will add value.
Market investment intentions provide context, not evidence of an individual programme’s return. An IAB Australia article reports that its survey found 53% of advertisers and agencies had increased overall affiliate and partnership marketing spend over the previous year, while 63% of advertisers indicated they would increase affiliate investment in the coming year.
Those figures do not establish the incremental value of any one programme.
Affiliate Marketing Investment Trends in Australia
- 2025% of advertisers increasing spend — 53%
- 2026% of advertisers planning to increase investment — 63%
Use reports as diagnostics
Review credited orders by publisher activity, customer status and placement. Awin Classic’s Publisher Incrementality report provides an always-on view of partner involvement across purchase initiation, basket values and conversion-rate impact, including an Initiator Ratio.
The report is available only to Awin Advanced customers. These observed measures can suggest what to investigate; they are not a randomised no-partner comparison.
Record the outcome, comparison, cost difference, uncertainty and conditions covered. Expand activity when evidence supports additional value at an acceptable cost. Change a weak placement or offer when a further test is justified, and revisit the result after material changes in promotion or customer mix.
The Awin report’s Basket Boost compares basket values when a partner is involved at any point in the journey with the advertiser’s site average order value over the same period. Awin’s example uses a £100 site average and £120 for orders involving a publisher, producing a 20% Basket Boost. This is a comparison with the site average, not a no-partner control group.
Awin defines Initiator Ratio as a measure of a partner’s ability to start new purchase journeys. Read it alongside the basket and conversion comparisons rather than treating any single indicator as proof of additional programme value.
Awin Publisher Incrementality Report Metrics
- Initiator RatioMeasures partner’s ability to start new purchase journeys
- Basket Boost20% increase in average basket value when publisher involved (e.g., £120 vs. £100 site average)
- Conversion Rate ImpactEffect on conversion rates across different touchpoints
In this guide
- Using a holdout to test an affiliate partnershipPlan a partnership holdout with controlled assignment, a shared store outcome, an observation window and checks for cross-exposure.
- Comparing new-customer rates across affiliate typesCalculate comparable new-customer rates across affiliate activities, show unknown status and avoid mistaking customer mix for lift.
- Estimating overlap between affiliate and paid searchCount observed affiliate and paid-search touchpoints on the same orders, show both overlap denominators and respect path-data limits.
- Deciding when an affiliate placement is not incrementalDecide whether a specific affiliate placement adds meaningful value using a defined outcome, credible comparison, uncertainty and cost.
