A delivery team books a six-week research phase before the build starts and plans nothing after it. The researcher will run the sessions alone and present findings at the end. Which two things does the Service Manual ask for that this plan misses?

The principle — two separate things go wrong when research is a phase run by one person: the timing and the audience.

Why the key is correct — on timing, the manual says this means This means doing small batches of user research in every iteration of each development phase - starting in discovery and continuing throughout live. A six-week block with nothing after it fails that on both counts. On audience, it says All team members should watch real users interacting with your service and talking about it - ideally for at least 2 hours every 6 weeks. The plan has one person in the room and everyone else reading a summary, which is a different activity with a different result.

Why the others are wrong — no amount of thoroughness reaches questions that have not arisen yet, and most of them arise once there is something to react to. A larger sample and a circulated report leave the shape untouched, since both failures are about when it happens and who attends. And adding a single phase after launch still leaves the entire build unresearched.

Remember this — every iteration, and everybody watches; a phase and a report is neither.

Sources — the UK Government Digital Service Manual on user research for government services.