Refill Kiosk Usability: When Stakeholders Disagree
MediLine, a pharmacy chain, is piloting a self-service prescription-refill kiosk in eight stores. The kiosk has a 15-inch touchscreen, a card scanner and a barcode scanner. The requirement document, written by the IT department, states that the kiosk must be “simple and secure”. In a usability test, 12 customers aged 65–82 attempted to refill a prescription without staff help. Only seven completed the task independently; five needed staff assistance. Four participants scanned their insurance card on the barcode scanner, and two abandoned after a timeout. The pharmacy manager wants two-factor authentication for privacy. The store manager wants the flow reduced to “three taps” to shorten queues. A nurse advisor insists on a confirmation screen with medication name and dose. No participant noticed the on-screen “Help” icon. The pilot ends in 12 days and the chain must decide whether to roll out to 60 stores. Write a design and research recommendation.
1. Separate stated requirements from observed user needs. What does the test data suggest about the real usability problems, and which findings are symptoms rather than root causes?
Read the worked answer
The stated requirements-'simple and secure'-are too abstract to test. The observed data point to three distinct problems. First, the card/barcode scanner affordance is ambiguous: four participants used the wrong scanner, which is a mapping error, not user carelessness. Second, the timeout and lack of visible help caused abandonment; the help icon failed because it was not noticed, so discoverability, not content, is the issue. Third, the nurse's confirmation screen is a safety requirement (medication name and dose), while the store manager's 'three taps' is a throughput goal and the pharmacy manager's two-factor authentication is a privacy control. These are not necessarily incompatible: a confirmation screen can be one step, and two-factor can be risk-based (e.g., only for controlled drugs or new devices). The root causes are unclear labeling, poor error recovery, and an unprioritized requirements conflict. A follow-up should test the revised flow, not just add features.
2. Propose a revised requirement set with at least three measurable acceptance criteria that reconcile the pharmacy manager, store manager and nurse advisor.
Read the worked answer
Rewrite the requirement set around user tasks and measurable outcomes. (1) Task success: at least 90% of adults aged 65+ complete a refill without staff help on the first attempt. (2) Error recovery: 100% of wrong-scanner scans trigger an immediate, specific message ('That looks like an insurance card; please use the card slot below') and allow retry within 10 seconds. (3) Safety confirmation: after scanning, the screen displays medication name, strength and dose, and the user must confirm with one tap; this replaces a separate 'three-tap' target. (4) Security: require a second factor only when the system detects a new device or a controlled medication, using a one-time code to the registered phone or staff-assisted verification; this satisfies privacy without adding friction to every transaction. (5) Help: a persistent 'Need help?' button with a staff call, not a small icon. These criteria reconcile the stakeholders because they make the trade-off explicit: speed is measured after safety and error recovery, not instead of them.
3. Design a short follow-up study (participants, method, metrics) that could be completed before the pilot decision.
Read the worked answer
Follow-up study: recruit 10–12 adults aged 65–85, including at least four who have low digital confidence and two who use a screen reader or magnifier. Use a within-subject comparison of the current kiosk and a clickable prototype of the revised flow. Tasks: refill a standard prescription, refill a controlled medication (simulated), recover from a wrong scan, and request help. Metrics: first-attempt task success, time on task, number and type of errors, whether the help button is noticed, and a single ease question. Also run a 30-minute stakeholder session with the pharmacy manager and store manager to agree the risk-based two-factor rule, and a short review with the nurse advisor on the confirmation screen wording. Because the pilot decision is in 12 days, use a high-fidelity prototype rather than a production build; the goal is to de-risk the flow, not to prove the final hardware.