Research guide
How to run a UX competitive analysis.
Choose a task. Collect the evidence. Compare how other products handle it. A practical method for your next UX review.
Make the task small enough to observe.
A UX competitive analysis compares how other products handle a task. Choose one you can follow from start to finish: creating an account, choosing a plan or setting up a first project.
Give each product the same assignment.
Pick a few competitors with a similar audience and task. Include an adjacent product if it solves the same interaction problem.
Keep the conditions consistent: a returning user on desktop may see a different experience from a new user on a phone.
| Task | Create a project and invite a teammate. |
|---|---|
| Starting point | The public homepage, using the same entry action. |
| Conditions | Record viewport, language, account state and the date. |
| Stopping point | Decide in advance: project created, invitation sent, or another observable milestone. |
| Access | Use only access you are authorized to have. Record steps you could not reach. |
Keep the screen behind the observation.
Capture the questions, choices and UI states that matter to the task. Keep the steps that led to each screen.
Write down what you can point to on the screen. “There is a skip option beside the invitation form” is checkable. “This onboarding is better” needs a criterion and evidence.
fossil captures competitor screens and flows from eligible web apps. Check for missing steps before comparing. This method also works with screenshots you collect yourself.
Ask the same questions of every flow.
Start with these criteria and adapt them to your research question.
| Criterion | What to observe | Evidence to keep |
|---|---|---|
| Information requested | Which questions or fields appear, and when? | Each relevant input screen. |
| Choices | What can someone choose, skip or defer? | The options and the paths reached. |
| Progress | How is the remaining work communicated? | Progress indicators and nearby copy. |
| Recovery | Can someone go back or correct a choice? | The observed action and resulting state. |
| Outcome | What does completing the task provide? | The final screen reached. |
Mark unreached steps as unknown. Record visible differences before discussing what they might mean for your audience.
Separate what you saw from what you’ll test.
For example, imagine a project tool that lets a new user postpone inviting teammates:
- ObservationThe invitation form has a visible “Skip for now” action.
- QuestionDo our users need to set up their project before bringing anyone else in?
- Next stepTest an optional invitation step with our users. Measure both project creation and later invitations.
Copy the comparison grid into your research document. Keep the captures, open questions and next steps beside it. A map of the app’s user flows helps teammates retrace the journey.