Design a requisition workflow that produces ready searches

Define the entry requirement

A manager starts with the outcome at risk, work currently uncovered, alternatives considered, expected timing, and proposed owner. This is enough to evaluate the need without demanding a finished job description too early. Separate request creation from approval so leaders can challenge whether hiring is the right response before recruiters polish a role that may never open.

Create meaningful workflow states

Use states that describe a real condition: draft, awaiting business approval, awaiting role design, ready for search, active, paused, filled, or closed. Define entry and exit criteria for each. Avoid a generic pending state that hides whether finance, the manager, or recruiting must act. Each state should identify one owner and the evidence required to move.

Make search readiness explicit

Before activation, confirm priority outcomes, essential evidence, level, reporting line, approved work setup, decision owner, interview capacity, and relevant budget approval. The recruiter checks coherence, not merely filled fields. If two inputs conflict, return the request with a named question and owner. Starting outreach while the role is unresolved transfers internal ambiguity to candidates.

• Business need and alternatives considered
• Approved role, level, and work setup
• Outcome-based brief and assessment owner
• Decision maker and interviewer readiness
• Current status, blocker, and next owner

Handle a returned request constructively

Illustrative example: a manager requests a senior analyst urgently, but the outcomes describe report production and the budget approval lists a different level. The workflow returns the item to role design with both mismatches visible. The manager and finance partner clarify decision ownership and approve the correct level. Recruiting begins two days later with one brief instead of discovering the conflict during an offer.

Review the workflow through exceptions

Track where requests wait, return, bypass steps, or change after activation. Sample cases to learn whether a field is unclear, an approver lacks capacity, or teams are using offline routes. Improve the mechanism rather than adding approval layers by reflex. For urgent exceptions, record who authorised the path, what evidence remains due, and which risk was consciously accepted. Give managers a visible view of their own requests and the next action without exposing sensitive information. Self-service status reduces chasing only when it names the blocker clearly. If people still ask for updates, review whether the workflow language describes internal states instead of the decision they are waiting on. Archive cancelled requisitions with a reason so they do not reappear later as apparently approved copies without context.