Time-Decay Attribution: credit decays as touches get further from conversion.

What time-decay attribution is

Time-decay attribution gives more conversion credit to touchpoints closer in time to the conversion. A touch 1 day before conversion gets more credit than a touch 14 days before. The decay rate is configurable. Google Analytics defaulted to a 7-day half-life (touches 7 days before conversion get half the credit of touches at conversion). The model fits short-consideration impulse purchases and direct-response campaigns where recent touches drove the action.

The model exists as a middle ground between last-touch (which gives 100 percent to the final touch) and linear (which gives equal credit to everything). Time-decay says recent touches matter more, but not exclusively. The result is a continuous gradient from old touches (low credit) to new touches (high credit).

How it works

Each touchpoint gets a weight based on its time-distance from the conversion. The standard formula is exponential decay: weight = 2 to the power of (negative time-since-touch divided by half-life). With a 7-day half-life, touches at conversion get weight 1.0, touches 7 days before get 0.5, touches 14 days before get 0.25.

Conversion credit is then proportional to weight. If three touches have weights 1.0, 0.5, and 0.25, total weight is 1.75. Credit shares are 57 percent, 29 percent, and 14 percent. Sum across the period to get channel-level performance.

When time-decay attribution fits

Time-decay is right for short-consideration purchases (DTC ecommerce, food delivery, impulse buys) and for direct-response campaigns where recent touches really do matter more. It is wrong for long-consideration purchases (B2B SaaS, enterprise software, real estate) where early-funnel research touches are often the most important.

The model also fits campaigns with seasonal or event-driven timing. A Black Friday campaign benefits from time-decay because the recent push-to-purchase touches matter more than awareness touches from October.

Common failure modes

Two failures. Using time-decay for long-consideration purchases where early research touches matter most. Picking a half-life that does not match the real consideration cycle.

Wrong cycle fit. A B2B SaaS purchase with 90-day consideration should not use a 7-day half-life. The model will systematically under-credit the awareness and consideration touches that drove the deal. The fix is to match half-life to the consideration cycle — 7 days for impulse, 14-30 days for considered DTC, 60-90 days for B2B.

Default half-life. Most teams accept the platform default (Google's 7-day half-life) without checking whether it matches their journey. The fix is to configure the half-life explicitly based on the typical time from first touch to conversion in your data.

Quick answers

What is time-decay attribution?

More credit to touches closer in time to the conversion. Touches further back get less. The decay is usually exponential with a half-life parameter.

What is the standard half-life?

Google Analytics defaulted to 7 days. For DTC and ecommerce, 7-14 days is typical. For B2B, 30-90 days is more appropriate.

When does time-decay fit?

Short-consideration purchases (DTC, food delivery, impulse), direct-response campaigns, and event-driven campaigns like Black Friday.

When is it wrong?

Long-consideration purchases like B2B SaaS or enterprise software where early research touches matter most. Time-decay under-credits those.

Should I change the half-life?

Yes if the default does not match your cycle. Configure based on real time-from-first-touch to conversion data.

How is time-decay different from position-based?

Time-decay is continuous; position-based is bucketed (40-20-40). Time-decay handles longer journeys more smoothly; position-based is simpler to explain.

Frequently asked

What is time-decay attribution?

An attribution model that gives more conversion credit to touchpoints closer in time to the conversion event. The standard formula is exponential decay with a half-life parameter.

How does time-decay work?

Each touchpoint gets a weight: weight = 2 to the power of (negative time-since-touch divided by half-life). With a 7-day half-life, touches at conversion get weight 1.0, touches 7 days before get 0.5, touches 14 days before get 0.25.

What is the standard half-life?

Google Analytics used 7 days as the default. For DTC ecommerce, 7-14 days is typical. For B2B with longer cycles, 30-90 days is more appropriate. Pick a half-life that matches the real consideration cycle in your data.

When is time-decay the right model?

Short-consideration purchases (DTC, food delivery, impulse buys). Direct-response campaigns. Event-driven seasonal campaigns where recent push-to-purchase touches matter more than older awareness touches.

When is time-decay wrong?

Long-consideration purchases (B2B SaaS, enterprise software, real estate) where early-funnel research touches are often the most important. Time-decay systematically under-credits those.

How is time-decay different from linear?

Linear gives equal credit to every touch. Time-decay weights recent touches more heavily. Time-decay produces more concentrated attribution to closing channels.

How is time-decay different from last-touch?

Last-touch gives 100 percent to the final touch only. Time-decay gives most credit to recent touches but still distributes some credit further back. Less extreme than last-touch.

Can I customize the half-life?

Yes. Configure it based on the typical time from first touch to conversion in your actual customer data. Default half-lives (often 7 days) frequently mis-fit longer-consideration journeys.