Reader and task analysis
4 min read
Lesson objectives
- Map audience knowledge, reading situation, and consequential errors.
Please sign in to use all the tools
Once you sign in, the reading tools are yours: larger or smaller text, typeface, line spacing, focus mode, highlighting and bold, and notes saved to your account and gathered in My study.
Naming a reader is more than choosing a job title. A receptionist using a procedure for the first time needs an entry point and field meanings. A supervisor intervening after failure needs escalation conditions and a record of what happened. They may work on the same service, yet blending both needs into one long sequence burdens the main path and hides the escalation decision. Start with a defined task and describe the reading situation: is the person learning before work, following steps during work, or recovering from an error? The order of information changes. A reader facing an error needs a diagnostic sign and a safe next action before a history of the system.
Make a useful reader profile with four elements: prior knowledge required by the page, immediate goal, available environment or permission, and consequence of error. Avoid inventing age, nationality, or educational level. “A new staff member with submission access but no approval authority” matters because it changes which instructions the page can give. If permissions are unknown, mark the profile as an assumption requiring confirmation. Do not target “everyone.” “Go to the administration panel” may be valid for a supervisor and impossible for a submission worker. Where two real paths exist, provide a clear decision point and separate steps, or a descriptive link to the supervisor path.
Convert the requester’s wishes into a reader-task matrix. For each row, record an observable action, starting point, required input, expected result, and a failure or handoff state. “Understand the system” is too broad; “check a request’s status by number and decide whether to wait or escalate” can be tested. Separate the primary task from a later secondary one. If 80% of a support team’s messages concern status, that does not prove 80% of all readers want that path; the message sample is biased toward people who needed support. Treat it as a reason to test status guidance, not as a fabricated audience proportion.
Prioritize rows using an explicit reason: frequency when supported, harm from failure, and dependency on earlier tasks. An uncommon step that prevents submission of wrong data may rank ahead of a common step that is easy to correct. Do not create a numerical score whose inputs cannot be defended. State why the order is chosen. Then run a quick path check: a new reader should identify the right path from heading and lead, while a supervisor should find the escalation trigger without traversing every basic step. If one cannot, the issue may be audience segmentation or document structure rather than a single word choice.
The product of analysis is an editorial decision, not an ornamental fictional persona. It identifies which path comes first, which concept must be explained, which action is blocked by permission, and when to link to another document. Keep only reader attributes that change those choices. Test assumptions with the process owner or a representative user when possible, and record the sampling limit. A desk review can still check paths and fields if no user is available, but it is not a real usability test.
Lesson summary
Map audience knowledge, reading situation, and consequential errors.