<b>Why a 90-day attribution window quietly underpays B2B partners</b>

<b>Why a 90-day attribution window quietly underpays B2B partners</b>

B2B sales cycles routinely outrun the cookie. Per multiple enterprise-software benchmarks, median sales-cycle length for deals above $25k sits in the 84–120 day range, with six-figure contracts often exceeding 180 days. A 30- or 90-day attribution window therefore truncates exactly the deals partners are best at: large, considered, multi-stakeholder purchases.

Three mechanisms drive the underpayment:

— <b>Window-cycle mismatch.</b> If the window closes before the median deal closes, the influenced revenue is structurally invisible. The partner sourced the account but gets zeroed.
— <b>Last-touch bias.</b> Short windows favor bottom-funnel re-engagement (retargeting, brand search) over top-funnel sourcing, where most partner value sits.
— <b>Survivorship in reporting.</b> Programs measure what closed inside the window and conclude partners drive few deals — a measurement artifact, not a behavior.

The corrective is not simply a longer window. Indefinite windows inflate fraud and over-credit incidental touches. A more defensible approach is to set the window to roughly the 75th-percentile sales-cycle length for the relevant deal tier, segmented by ACV band, and to pair it with a multi-touch model for any deal where a partner appears in the sourcing position.

<b>Implications:</b> attribution windows should be a function of deal economics, not a platform default. Audit your window against your own closed-won cohort before concluding partners underperform — the two are easily confused.

<b>Open question:</b> when first-touch partner-sourcing and last-touch self-serve conflict, what credit split survives a finance audit without incentivizing either gaming?