Short vs. long attribution windows: tuning to your sales cycle

Short vs. long attribution windows: tuning to your sales cycle

The attribution window — how long after a partner touch you'll still credit the conversion — is a single number that silently decides which partners get paid. Set it from your actual cycle data, not a B2C default.

Short windows (30-90 days) inherited from B2C.
— Crisp, recent, easy to defend; the platform default almost everywhere.
— In B2B, often catastrophically short. If your median sales cycle is two-plus quarters (common in mid-market and enterprise per repeated cycle-length benchmarks), a 30-day window credits almost no partner who works the top of the funnel. The awareness partner introduced the account in January; the window expired in February; the deal closed in August crediting no one — or the nearest bottom-funnel partner by default.

Long windows (180-365 days) matched to enterprise cycles.
— Capture the full influence path; credit early-stage sourcing partners fairly.
— Weakness: they inflate the influence claims of stale touches and invite credit for partners whose contribution decayed months ago. A 12-month window will credit a podcast someone half-remembers.

The disciplined approach: set the window to your measured cycle distribution — typically the 75th-90th percentile of time-to-close — not a round number and not the platform default. And recognize the window interacts with your model: a long window plus last-touch still starves top-funnel partners; the window only helps if your model distributes credit across time.

Trade-off: Short windows are clean but amputate long-cycle influence. Long windows capture it but credit stale, low-influence touches.

Open question: Is your attribution window a measured parameter, or just the platform's B2C default nobody changed?