The People & AI glossary.

Find a term. Open the explanation. Keep your place.

5 terms

Assumption testA small check of a belief that a proposed solution depends on.

Read the explanation

State what needs to be true, how you will check it and what finding would change your decision.

Test the consequential uncertainty before building the whole solution. A successful test of one belief does not establish the full business impact.

Fictional agency: before assigning a deputy approver, check whether that person has the authority and capacity to make the decision.

Checking that a deputy can approve is an assumption test. Measuring campaign delivery after a rollout answers a different, broader question.

Assumption Testing: Everything You Need to Know to Get Started

Continuous discoveryA recurring practice of learning from people’s experiences and checking ideas while you work towards an outcome.

Read the explanation

This People-work adaptation draws on Teresa Torres’s product discovery approach. Employee conversations and small tests inform what you build next.

Keep returning to recent experiences, including those that challenge your explanation. Use a proportionate rhythm for the decision and its risks.

Fictional agency: after changing handovers, speak with the next people who use them and update the opportunity map from what happened.

It is ongoing learning tied to decisions. It is not repeated surveying without a response or a reason to postpone necessary action.

OpportunityA need, pain point or desire worth addressing because it may help you reach an outcome.

Read the explanation

Describe the person’s experience in its context. Keep the possible response separate so you can compare more than one way to help.

An interview can suggest an opportunity. Check it against further accounts and records before treating it as the explanation for everyone.

Fictional agency: a designer needs to know which version of a brief has been approved before committing time to it.

“Know which version is approved” is a need. “Install an AI approval bot” is one proposed solution.

Opportunity Solution Trees: Visualize Your Discovery to Stay Aligned and Drive Outcomes

Opportunity solution treeA visual map connecting an outcome to needs, possible solutions and the assumptions you need to test.

Read the explanation

Teresa Torres developed this tool for product discovery. Here it is adapted to employees’ experiences of work.

Branches keep alternatives visible. They make the reasoning inspectable without claiming that every connection has been proven.

Fictional agency: fewer stalled handovers → know who can approve → named deputy, shared approval window or routing reminder → check which assumptions hold.

It is a map of possible paths and learning. It is not a backlog of solutions already chosen for delivery.

Opportunity Solution Trees: Visualize Your Discovery to Stay Aligned and Drive Outcomes

OutcomeA change in people’s experience or business performance that you want to observe.

Read the explanation

Name the change before selecting the thing to build. Give it a scope and a way to notice progress.

In People work, include who should benefit and what must not get worse. A faster handover is less useful if it creates unmanageable work for the approver.

Fictional agency: fewer campaign handovers stall while people work out who can decide.

A new approval checklist is an output. Less time waiting for a decision is an outcome.