July flash sale
Demand management determines which requests deserve discovery, which options are ready for commitment, and which ideas should be delayed or discarded. A Demand Manager improves the upstream system so delivery capacity receives a purposeful mix of work instead of an uncontrolled stream of stakeholder requests.
Official reference: Kanban University Demand Manager course. Course names, prerequisites, credentials, platform features, and award rules can change; confirm them on the official page for the class or credential you are considering.
Build a complete demand picture
Capture request source, customer purpose, work type, arrival date, urgency rationale, expected outcome, risk, and current decision state. Include rejected, abandoned, duplicated, and withdrawn requests. Measuring only committed delivery work hides the scale and shape of demand and prevents leaders from learning why customers leave or why the upstream queue grows.
Separate ideas, options, and commitments
An idea can be recorded with little investment. An option has enough evidence to support a future decision. A commitment consumes constrained delivery capability and carries an expectation. Visualize these states and limit discovery work. Without explicit boundaries, every suggestion becomes a promise and product teams spend their time refining options that may never be selected.
A working-service example: the Demand Manager role
A product group maintains 480 backlog items and still reacts to executive messages. The Demand Manager introduces an idea pool with a ninety-day review, limits discovery to six options, and requires a customer purpose plus a measurable fitness criterion. Replenishment selects only options that fit available delivery capacity. The backlog becomes smaller because weak options are intentionally retired, not because somebody performs a cosmetic cleanup.
Use decision filters rather than stakeholder volume
Define filters connected to customer purpose, strategic fit, urgency, cost of delay, evidence strength, risk, and effort to learn. State who owns each decision and how conflicts are resolved. A filter is not an automatic scoring machine; it structures judgment and makes inconsistent treatment visible.
Connect upstream and downstream capability
Replenishment should consider delivery capacity, work-type mix, dependencies, and aging committed work. Pulling ten large options into a constrained system does not create value sooner. Use historical throughput and current WIP to constrain selection, and keep unselected options visible without pretending they are scheduled.
Test product-market and service fitness
Ask which customer purpose an option serves and which fitness criterion it changes. Use interviews, small experiments, prototypes, and service data to reduce uncertainty. Discovery should terminate weak options as well as strengthen good ones. Track decision time and abandoned demand alongside feature throughput.
Map the decision with this the Demand Manager role template
| Demand state | Minimum evidence | Decision | WIP policy |
|---|---|---|---|
| Idea | Purpose and source | Record or reject | Large pool; age review |
| Discovery option | Risk and learning question | Test or defer | Strict discovery limit |
| Ready option | Outcome, dependency, acceptance | Select or retain | Limit ready inventory |
| Committed | Capacity and expectation | Pull into delivery | Delivery-system WIP limit |
Use the Demand Manager role in a real service conversation
- Record all intake channels and discarded demand.
- Define idea, option, ready, and committed states.
- Limit concurrent discovery.
- Publish decision filters and decision owners.
- Use capability evidence during replenishment.
- Review customer-purpose evidence after delivery.
Introduce the Demand Manager role without a big-bang change
During the first week, use Demand Manager Playbook for Upstream Kanban and Discovery to establish a shared boundary and baseline. Begin with this action: Record all intake channels and discarded demand. Invite the people who request, perform, manage, and receive the work; their different views will reveal assumptions that a board or dashboard cannot settle alone. Record definitions, missing data, known exceptions, and current customer consequences. Do not change several policies during the baseline week, because the service needs a credible comparison for the experiment that follows.
During weeks two and three, complete these actions: Define idea, option, ready, and committed states. Limit concurrent discovery. Publish decision filters and decision owners. Select one policy experiment that is within the group's authority, state why it should influence the observed behavior, and name a safety boundary for quality, workload, compliance, or customer harm. Keep unrelated changes visible. Use the working table above during the relevant cadence so the resource becomes part of a decision rather than a document that people read once.
During week four, complete the remaining actions: Use capability evidence during replenishment. Review customer-purpose evidence after delivery. Compare the new evidence with the baseline, ask affected customers and service participants what changed, and look for displaced delay outside the original boundary. Decide explicitly to keep, adapt, stop, or extend the experiment. Store the decision beside the policy and link back to the Kanban University Demand Manager course, noting the access date, so future reviewers can distinguish official guidance from the local interpretation used in this service.
Training options for deeper the Demand Manager role practice
Develop the relevant foundations through Kanban System Design training, Kanban Systems Improvement training. Continue with upstream product-discovery guide, demand-analysis guide. Choose the path that matches your service responsibility and apply the learning with the people who operate and use the service.
What to ask after trying the Demand Manager role
- Which customer or service decision should this Demand Manager Kanban playbook help us make?
- What evidence do we have, and where are the measurement boundaries unclear?
- Which policy or behavior is within our authority to change?
- What unintended consequence should we watch during the experiment?
- When will we review the outcome and decide to keep, adapt, or stop?
