Professional Business Analysts / Finding out

Watch the work before asking about it

What people describe and what they do differ in useful ways.

An hour spent sitting beside somebody doing their job produces information that six interviews will not. Not because interviewees are dishonest, but because expertise becomes invisible to the person who has it. The steps that are automatic do not get mentioned. The exceptions handled forty times a day do not feel like exceptions. The second system that is opened for one field is not part of the process as they understand it.

Observation surfaces exactly those things. It also surfaces the workarounds, which are the highest value findings available, because each one marks a place where the official process does not fit the actual work. A person who maintains a private spreadsheet alongside the system of record is telling you something specific about what the system fails to do.

The technique is unremarkable and requires only discipline. Sit where you can see, ask nothing for the first stretch, note what happens rather than what you expect, and save the questions for afterwards. The questions are better for having waited, because they are about something you saw rather than something you imagined.

The trap is being seen as an evaluator. If the person believes their performance is being assessed, they will perform the official process, which is precisely the version you already have in a document.

There is one further discipline that improves observation considerably, which is to note the interruptions. Most process documentation describes an uninterrupted task, and almost no real task is uninterrupted. What a person does when the phone rings partway through, where they leave the work, and how they resume it, determines a great deal about what a system needs to support, and none of it appears in a description that begins with the first step and ends with the last.