Cost of Delay: the real price of postponing
Postponing a delivery has a price, and it runs per week, even when nobody puts it on the spreadsheet. Order the features below by different policies and see which sequence loses the least money.
Four features compete for one sequential team (one at a time). Each has a Cost of Delay (what you lose per week while it is not delivered) and a duration. In which order do you deliver them to lose the least?
Your sequence
You ordered by highest value per week (CoD). Total: $900, or $100 (13%) above the $800 minimum. The most valuable item is large and, coming first, pushed everything else back.
Value alone misleads: an expensive but long item makes the whole queue wait. What matters is cost per week divided by duration (CD3/WSJF), not cost per week alone.
It is the "do the most rewarding thing first" trap: if that item is slow, the next weeks of returns are stuck behind it. The fix is weighing value against time.
Before pulling the most valuable item, a pro asks how long it blocks the others. If it is long, an almost-as-valuable but much shorter item should jump the queue.
Urgency profiles: value lost as delay grows
The golden rule
Sequence by CD3 = CoD ÷ Duration (the WSJF logic), not by value alone (CoD) nor by shortest size (SJF). Doing first what has a high cost per week AND is quick to finish minimizes total loss.
What cost of delay is
Cost of Delay (CoD) is how much you lose per unit of time while a delivery has not happened: revenue not captured, risk that grows, opportunity that passes. It is the missing link between "value" and "urgency": it joins how much something is worth to how time-sensitive it is. Without CoD, prioritization is opinion; with CoD, it is arithmetic.
CD3 / WSJF: the order that minimizes the loss
CD3 = Cost of Delay ÷ Duration
When a team delivers one thing at a time, every feature waiting in the queue keeps accruing cost of delay. The question is not "what is worth more?" (CoD alone) nor "what is faster?" (SJF), but "what gives the most value per week of work?". That ratio is CD3 (Cost of Delay Divided by Duration), the very logic of WSJF.
Sequencing by descending CD3 mathematically minimizes the total cost of delay incurred: high cost-per-week, short-duration deliveries go first, clearing the queue and cutting the accumulated loss. In the lab above, the CD3 order always beats CoD and SJF. Try reordering and confirm it.
Urgency profiles
Not every cost of delay grows the same way. A linear profile loses value at a steady rate; a fixed-date one barely costs anything until a deadline, then it drops off a "cliff"; an urgent one loses a lot of value right at the start and then levels off; an intangible one looks cheap for weeks until the loss spikes late.
Knowing the profile changes the decision: fixed-date items should be timed to the deadline, urgent ones go to the top of the queue, and intangible ones cannot be postponed forever just because they "do not hurt" today. The profiles chart in the lab shows the four curves side by side.
Frequently asked questions about cost of delay
What is Cost of Delay?
It is how much you lose per unit of time while a delivery is not finished: revenue, opportunity or risk. It connects value and urgency in a single number.
What is CD3 and how does it relate to WSJF?
CD3 is Cost of Delay Divided by Duration. It is the same logic as WSJF: prioritize what delivers the most value per week of work.
Why not prioritize by highest value (CoD) alone?
Because a very valuable but long feature clogs the queue and makes several others accrue cost of delay. Dividing by size (CD3) minimizes the total loss.
What are urgency profiles?
They are the shapes in which value is lost over time: linear, fixed-date (cliff), urgent (steep early loss) and intangible (loss that spikes late). The profile changes the best moment to deliver.
Enjoyed learning by doing?
Take prioritization to the next level: manage a squad in the Kanban Game.
