1️⃣Research guide
👋🏼 Introduction
3 steps is all it takes to do research. Do it as often as you can. It’s the best way to build your intuition for the product space, for consumer sensibility, and conviction in your designs.
Get started with the Research Template [Duplicate me]!
1️⃣ Build hypothesis
Before you start your research, it’s important to answer what you are testing and how you plan to test it.
In the Experiment table, fill the first three columns.
Identify a hypotheses: These are assumptions in your proposed design which you are testing. Be clear and opinionated about your hypothesis to get the clearest reaction to your designs.
Create an experiment: Write down the specifics of the task, the hypothetical situation, or the question you’ll ask to test each hypothesis. Be open ended in the phrasing here. “How, what, why…” type questions rather than “Would you do x or Y…” type questions.
Define success criteria: This helps you be clear about what counts as successful performance of the design. It could be a click, it could be the time taken on the screen, it could be selection of the specific option based on what you’re testing
2️⃣ Test with users
Recruit users: If you have a user base already, that’s an easy place to recruit from. Make sure you filter for the kind of users you are likely to use the feature you’re testing and are active users of the product if that’s important. For additional help with recruitment, refer How to recruit users and Recruitment Script.
Prepare an interview guide: Use the Interview Guide template to identify your warm-up and wrap up questions to put participants at ease.
Take notes during interviews: Take notes yourself or with the help of a note-taker. Observe behaviour, comprehension, and inferences. Note these observations down rather than taking verbatim notes.
In the Experiment Table, fill the “Notes” and “Results” columns for each participant:
Determine results: Mark overall success / failure of each hypothesis with each participant. As a rule of thumb, if it’s not a strong yes, it’s a no. This default will help avoid mediocre designs and help the team take the harder decisions necessary to build performant products.
3️⃣ Synthesise Insights
Write your insights: For each hypothesis, identify number of success / failure across participants. Go through the notes you’ve taken and identify reasons or patterns for the result. Continue this step for all hypotheses to gather all the insights from the study.
Create recommendations: Once insights are in, step back and identify the unsolved aspects of the designs, or new Jobs to Be Done (JTBD) that might have surfaced. It might even be aspects of designs to retain as is. Stay away from ideating at this point.
Assign next steps: The study is incomplete until each recommendation is either assigned an owner or is clearly marked as not important to act on at this time.
In the Experiment Table, fill the last three columns.
Last updated
