Cost of Delay & WSJF Prioritization Guide

The Cost of Delay formula

CoD ($/week) = Business Value ($/week) + Time Criticality ($/week) + Risk Reduction / Opportunity Enablement ($/week). Business Value is the revenue or savings the feature unlocks per week once live. Time Criticality captures decay — a competitor launching, a contract deadline, or a seasonal window that shrinks the prize the longer you wait. RRO&E is the value of options the feature opens (a platform capability that unblocks two future bets). Sum the three and you have a single defendable dollar-per-week number.

Why senior PMs use CoD instead of RICE

RICE, ICE, and MoSCoW all output abstract scores. The moment a CFO asks 'what does this cost us per quarter of delay?', those frameworks run out. CoD is dollar-denominated, which lines up directly against fully-loaded squad cost and the discount rate finance already uses. When you argue for headcount, a platform bet, or an at-risk feature slip, CoD gives you the economic case in the language the board approves budgets in.

WSJF: turning CoD into a ranked queue

WSJF = Cost of Delay ÷ Job Size (or Duration). Two features with the same CoD are not equal — the smaller one delivers the same weekly value in less calendar time, so shipping it first buys more total value across the queue. WSJF is provably optimal for maximizing economic throughput on a fixed-capacity team: rank by WSJF descending and ship top-down. It is the default sequencing method in SAFe for exactly this reason.

Worked example: two features on one squad

Feature A — checkout redesign. Business Value $18K/wk (measured from a similar redesign), Time Criticality $6K/wk (competitor ships in Q3), RRO&E $2K/wk. CoD = $26K/wk. Job Size = 8 weeks. WSJF = 3.25. Feature B — internal admin tool. Business Value $9K/wk (support-time savings), Time Criticality $1K/wk, RRO&E $4K/wk (unblocks a compliance workstream). CoD = $14K/wk. Job Size = 3 weeks. WSJF = 4.67. Feature B ships first — smaller job, higher weekly value per week of team time. Total value across both jobs: 208K vs. 42K if you'd shipped A first and B slipped 8 weeks.

CoD, NPV, and the horizon question

CoD answers 'what does another week of delay cost us?'. NPV answers 'is this feature worth building at all?'. Use CoD to sequence the queue once you have a shortlist of positive-NPV features; use NPV to decide whether to keep the feature in the shortlist. On Road map roi, the NPV calculator prices the total case and CoD/WSJF sequences the ranked queue on top — same discount rate, same cash-flow inputs, one memo.

Where CoD breaks (and how to defend the number)

The failure mode is fabricated confidence: teams pull a Business Value number from the air, call it CoD, and defend it as if it were measured. The discipline is the same as NPV — anchor Business Value in a specific measurable (activation lift, GMV, support tickets deflected), apply the same confidence haircut RICE uses, and re-price CoD monthly as the underlying data changes. A CoD that never updates is a stale opinion, not an economic argument.

Cost of Delay vs. other prioritization frameworks