Optimizing Business Value
What this domain tests
Optimizing Business Value examines how value is defined, prioritized, delivered, and measured. Questions cover outcomes versus outputs, value streams, cost of delay, weighted shortest job first (WSJF), minimum viable product (MVP) thinking, feedback loops, and benefits realization.
Core rules and concepts
- Value is a benefit to the customer or business, not the number of features delivered. Output metrics can rise while value falls.
- Prioritization should make economic trade-offs explicit. Cost of delay expresses the loss from not delivering sooner; WSJF compares cost of delay to job size or duration.
- Small batches and experiments reduce risk and accelerate learning. An MVP is the smallest thing that tests a hypothesis, not a half-built product.
- A value-stream view optimizes flow end-to-end. Local efficiency in one team or department can worsen overall lead time.
- Define success measures before delivery and track outcomes and benefits, not velocity alone.
- Customer collaboration and fast feedback turn assumptions into evidence.
Worked example
A product owner compares two backlog items. Item A has a cost of delay of 200,000 and a job size of 4 weeks. Item B has a cost of delay of 90,000 and a job size of 1 week. Using WSJF, A scores 50,000 per week and B scores 90,000 per week.
Reasoning: B has the higher WSJF, so it should generally be sequenced first if capacity and dependencies allow. The decision is not based on the larger total benefit of A; it is based on delay cost per unit of effort. The product owner should also check whether B unlocks learning or reduces risk that changes A's value.
Traps that cost marks
- Prioritizing by the loudest stakeholder or the largest visible feature.
- Using velocity as a value metric; velocity measures capacity, not benefit.
- Ignoring cost of delay and treating all backlog items as equally urgent.
- Building a full feature when a smaller experiment would test the assumption.
- Optimizing one team's utilization while increasing end-to-end wait time.