Outcome-based buying

Outcome-based buying is purchasing media against a business result, such as a sale, a lead, an app install or measured lift, with the price, guarantee or goal tied to that result rather than to impressions.

There are three main forms. In the lightest, the buyer still pays per thousand impressions but the platform optimises toward an outcome. In the middle, the buyer pays per outcome, as in cost per acquisition. In the strongest, the seller guarantees an outcome and makes good if it misses. The CTV platform tvScientific launched cost-per-outcome buying on 17 August 2023 (tvScientific, 17 August 2023). Buyers pay for a predefined result, such as sales or post-campaign site traffic. The company calls cost per outcome broader than cost per acquisition, because one outcome can bundle several actions.

Outcome deals shift risk to the seller, who then wants control over targeting and placement. Buyers lose some transparency in return. Both sides must agree in advance on the measurement: the attribution window, the source of sales data, the treatment of invalid traffic and, ideally, a test of incrementality.

Example: Amazon Ads launched Full-Funnel Campaigns on 29 September 2026 (Amazon Ads, 29 September 2026). The advertiser supplies budget, products and creative. Amazon's AI then plans and optimises sponsored ads, display, video and streaming TV in one campaign, and reports new-to-brand and long-term sales.

Related terms: CPA, attribution window, incrementality, ROAS, new-to-brand, CTV.

Authoritative source: tvScientific, "tvScientific Launches Cost-Per-Outcome CTV Ad Buying Through Partners" (17 August 2023, company release).