Linear vs. time-decay multi-touch: which weighting fits your sales motion

Linear vs. time-decay multi-touch: which weighting fits your sales motion

Once a program graduates from single-touch, the next fork is how to distribute credit across touches. Linear (equal credit to all) and time-decay (more credit to recent touches) are the two most-deployed multi-touch models, and they encode opposite assumptions about how B2B buyers actually decide.

Linear assumes every interaction matters equally.
— Useful when the buying committee is large and influence is genuinely diffuse — committees of 6-10 stakeholders are common in enterprise deals per 2023 buyer-group research.
— Weakness: it dilutes credit so thinly that no single partner sees a payout worth optimizing toward.

Time-decay assumes proximity to close signals influence.
— Fits velocity-driven, single-champion deals where late momentum is real.
— Weakness: it structurally reproduces last-touch bias, just softened. Awareness partners still lose.

The under-discussed variable is your half-life setting. A 7-day decay half-life on a 9-month cycle is effectively last-touch wearing a multi-touch costume. Match the half-life to median cycle length, or the model lies.

A position-based (U-shaped) model — heavy weight on first and last, the rest split — is the pragmatic compromise when you want to honor both sourcing and closing without modeling every touch. It is harder to explain to partners, which is its real cost.

Trade-off summary: Linear is fair-feeling but motivationally inert. Time-decay is motivating but quietly biased. Position-based is balanced but opaque.

Implications: Pick the model whose bias you can defend to the partner it disadvantages — because that conversation will happen.