5 steps to write an effective discussion guide
Struggling to get started with your next research project? Whether you are new to UX research or just exploring a new topic or team, it can be difficult to transform what you plan to do into an actual plan of action — your discussion guide.
What I’ll share worked for me for in-depth interviews, prototype testing and foundational research for specific digital product areas. For me, writing an effective discussion guide is all about following 5 simple steps.
#1: Define goals, research questions and hypotheses with stakeholders.
Start with the foundations of your research project:
- Agree on the goal(s) and research questions of the project. What do we want to learn? What will we actually do with the insights we gather? What problem will we solve for our users? Make sure to challenge goals that you feel are not realistic or not contributing to solve problems for your users.
- Form hypotheses to test. If you form testable hypotheses to test in the research, it helps you to write the discussion guide in a structured way. It allows you to generate scenarios to test, provide feedback on realistic designs to include and it forms a base for your analysis later on. Your report will then have tangible results in relation to the hypotheses you tested, which is great if you’re planning to test more prototypes over time. But if you are conducting foundational research, I’d argue testable hypotheses are equally important — it helps you create a library of work over time and it gives you focus on what you set out to learn, rather than getting sidetracked by new questions.
If you ever underinvested in defining goals and hypothesis, you know you can’t really over-invest in this step of your research project. 😉 Here’s a great read on writing a research plan by Margie Mateo Villanueva.
#2: Structure your guide into key sections, each serving a unique purpose
Unique sections allow you to focus and check if you’re covering everything you set out to learn in your goals. As an example, here’s a common structure I follow:
- Introduction & Context: to learn from which ‘context’ participants share insights and to build rapport early in the conversation.
- Deep-dive into a specific product area: related to your research questions and goals, followed by the prototype (if you’re testing one)
- Foundational deep-dive: whether or not I’m conducting foundational research, I always try to include some foundational questions in evaluative research (see point 5). This is a great place to include creative exercises and end the session on a high.
- Conclude the interview with questions: from your audience (if you have them) and from the participant. Depending on the area of business you’re in, participants may participate in the research with the underlying motivation to learn something (especially in B2B research) — which is why it is crucial to truly leave time for questions.
(Want to know more? Gina Knox published a great note on discussion guides for prototype testing).
#3: Trial and error
Not all methods and questions work for every situation or project. But if you don’t continue to try out new ways to generate insights, your discussion guides and research might turn into outdated artefacts that don’t generate the insights you’re after. I always try to make it a point to include a new question, which is how I got some of my current favourite questions (often borrowed from other researchers). Examples include:
- ‘How would you describe what you’ve just seen (product, prototype) to a friend or colleague?’
- Ask participants to bring a photo that represents something related to the interview topic, for example a picture that describes how they feel when they use your product
- End with an exercise to write a pro and con list about the product/prototype (a ‘B2B’ friendly version of the ‘break-up letter’ used in research)
Besides the insights they create, it also gives me a boost of energy if new questions turn out to be surprisingly effective. 💁
#4: Ask for feedback, but carefully consider from whom and on what
I realised early on that feedback is a great tool to improve your research efforts, but it can also open the floodgates to discussion and feedback on things you’re not looking to amend. For example, non-researchers who discuss the best ways to ask questions, while you carefully considered avoiding bias to generate deep and objective insights. Instead, I generally ask for more specific feedback:
- From non-researchers: Feedback on what you want to learn, instead of how you want to get to insights
- From fellow researchers: Feedback on how to get the insights
- From anyone involved in the project: Feedback that you read the document and confirm it aligns with the goals agreed on
At first, it might sound difficult to restrict people in the feedback they give. But in my experience people are able to give higher quality feedback when you specify what it is you’re after.
# 5: Create room for insights of the future
In many organisations there might not be time or budget for dedicated foundational research. However, that doesn’t mean you can’t prove the value of it through evaluative research.
But.. How? 🤔
By using the time and effort invested to organise a prototype test or evaluative study. I usually add questions that are related to the topic in question, because I think they will become relevant in the future. It depends on the organisation you’re in, but I personally haven’t experienced a lot of push back on this. Over time, you can also refer to a case study you make by describing an opportunity that was pursued through the simple road of adding in questions to existing research. Some of these explorations don’t result in actual opportunities to pursue, and that’s ok. Because when they do, it’s an amazing way to become a crucial strategic partner and leader in a product area and to see your efforts grow into something real! 🌱
These simple steps helped me from the start of my career all the way through to when switching to a new domain recently. Don’t be a stranger, let me know in the comments what has helped you! 👋
