Cost of Delay answers one question: what do we lose if we wait?
Most prioritization stalls between whoever is loudest and whoever has the strongest gut feeling. Cost of Delay asks a sharper question that the whole team can reason about: what actually gets more expensive the longer it sits unfinished?
Let the class of service describe the shape of the cost
Different work loses value in different shapes over time. ScrumDo ties each class of service to a cost-of-delay curve, so you pick the shape that matches reality instead of scoring everything the same.
| Class of service | What it means | Cost-of-delay shape |
|---|---|---|
| Standard | The default, pull in order | Rises, then levels off |
| Fixed delivery date | A hard date drives it | Steady, then a sharp cliff on the date |
| Expedite | Drop almost everything, one at a time | High and steep from right now |
| Intangible | Slack-fill; escalate when it bites | Low now, can spike much later |
Assigning a class gives a card its economics for free. You only override the defaults when you have better evidence.
What economic priority means
ScrumDo turns the curve into a single number you can sequence by, so very different releases compare on the same footing. The principle is the well-known one from Lean product economics:
- Start from the Cost of Delay, the value lost while the work waits.
- Express it as a rate: Cost of Delay per week of delay.
- Weigh it against effort by dividing by how long the work takes to build.
The result is economic priority. Higher means finishing it sooner protects more value per unit of effort.
A worked example a finance team will recognize
A regulator sets a filing deadline. The compliance release protects about $60,000 a week in premium while it is on time; miss the date and value collapses to a $5,000-a-week residual (fines and market-exit cost are tracked separately). The team is 4 weeks behind and the build takes 8 weeks. That is a Fixed delivery date class.
| Input | Value |
|---|---|
| Class of service | Fixed delivery date |
| Peak value protected | $60,000 / week |
| Deadline | Week 12 |
| Residual after the date | $5,000 / week |
| Weeks of delay | 4 |
| Weeks to build | 8 |
| Output | Value |
|---|---|
| Total Cost of Delay | $20,000 |
| Cost of Delay per week | $5,000 |
| Economic priority | 625 |
Before the deadline the cost of waiting is low; miss it and you lose the $5,000/week residual, so four weeks late costs $20,000. Economic priority = Cost of Delay per week ÷ weeks to build = $5,000 ÷ 8 = 625. That one number lets leaders sequence this release against everything else on the board, with the assumption note and confidence attached so the math can be questioned later.
Sequence by class first, then break ties
Class of service sets the outer order: Expedite first, then Fixed delivery date, then Standard, then Intangible. Within a class, economic priority sequences the work by what waiting actually costs. And within a single release, source-card WSJF, business value, time criticality, risk reduction, and size, only breaks the remaining ties. It never overrides class policy or the portfolio economics.
Key terms
- Cost of Delay
- The value lost for each week a piece of work waits unfinished.
- Economic priority
- Cost of Delay per week divided by the weeks it takes to build; higher means sequence it sooner.
- Class of service
- A policy that sets how work is treated and the shape of its cost-of-delay curve (Standard, Fixed delivery date, Expedite, Intangible).
- Confidence
- A signal to the next reviewer about how grounded the inputs are, not a statistical confidence interval.