CPL, CPA or revenue share: which performance model fits your campaign?

With cost per lead, the buyer pays for a contact enquiry meeting agreed criteria. The model provides fast feedback and works when audience, region, data, intent, exclusivity and duplicate rules can be checked objectively.

Risk after delivery largely sits with the party handling follow-up. The CPL therefore needs room for contactability and sales conversion. A cheap lead is not economical when the validity threshold is so low that sales processes large volumes of unsuitable enquiries.

Cost per acquisition or action pays for a later stage, such as an appointment, accepted application or sale. This can align more closely with commercial value, but only when the event is measured consistently and returned quickly.

The later the measurement point, the more factors outside the traffic partner's control affect it. Sales capacity, price, availability and response time can change the result. A CPA model becomes unsustainable when a supplier carries all risk but has no visibility or influence over the final stages.

With revenue share, a partner receives an agreed portion of revenue or margin. The model can align incentives strongly, especially with recurring revenue or long-term customer value. It also requires the most transparency around transactions, cancellations, refunds, costs and attribution.

Define whether the share is based on gross revenue, net revenue or profit. Agree which costs count, how long a customer remains attributed and when figures become final. Without these definitions, disagreement moves from lead quality to accounting interpretation.

A later payment point usually creates more time between media expenditure and payment. The party financing spend also carries working-capital risk. Fast, reliable feedback can make this manageable; monthly manual reports usually cannot.

Assess each model across five points: objective measurability, feedback speed, influence of each party, pre-financing and room to scale. A model can be theoretically fair and still fail operationally.

A new campaign does not yet have a stable conversion chain. A test budget, fixed production contribution or CPL with a bonus for later quality can distribute risk while both parties gather data. The model can change after enough mature conversions.

Define the evidence required for the next phase before launch. This prevents a pilot from becoming a permanent exception and avoids promising full performance before market, funnel and follow-up have been proven.