When the feedback starts coming in, you're really only ever asking one of two questions. Either "what did people make of this step?" or "what did this particular tester do?"
Those two questions want the data laid out in opposite directions, which is why the results page has two views and a toggle between them. Pick the right one and a job that takes an hour of scrolling takes a couple of minutes.
In this article
- Switching between the views
- By step: what did people think of this?
- Table view: what did this person do?
- Shaping the table
Switching between the views
The toggle sits on the Feedback tab of your results page, with two options: By step and Table view. All it changes is the layout. Nothing gets filtered out when you switch, so you can flip between them as much as you like while you're chasing an answer.
By step: what did people think of this?
By step opens on your tasks, each one showing how many steps it has and how many testers made it through. Open a task and you get its steps with everyone's responses underneath: one step, all the answers to it, then the next step.
The By step view listing tasks with per-task completion
Look at that completion figure before you go any further. If only half your testers reached a task, that changes how much the answers inside it are worth, and it's not something you'd guess from reading the answers themselves.
This is the view for judging your product. If you want to know whether your onboarding confused people, you want all the onboarding answers in one place and you don't much care who said what. It's also much better for free-text feedback, where reading twenty answers to the same question one after another shows you a pattern that no single answer would.
Expand all and Collapse all help you move around quickly, and Export all pulls everything, timestamps included, into a spreadsheet.
Table view: what did this person do?
Table view pivots the same data: a row per participant, a column per step, with the columns grouped under the task they came from. Now you're reading across rather than down.
This is the view for judging your testers, and for spotting the shape of the data. A row with three answers and eight blanks is somebody who dropped off part way through. A column where everyone gave the same answer is a step that isn't telling you anything. You won't see either of those reading step by step.
It's also the only way to cross-reference, since one tester's answer to step 2 sits next to their answer to step 9 instead of being two sections apart.
The table view with one row per participant and columns grouped by task
Shaping the table
The table is built to be cut down.
Search filters the rows to whatever matches. Columns opens a panel for showing and hiding individual columns, so you can strip the whole thing back to the two steps you're actually interested in. And Export view downloads what's left on screen, visible columns and filtered rows only, which is covered in Export Your Results and Tester Data.
Your column choices stick per beta, so the layout you set up for one project is waiting for you next time, and another project keeps its own.