Research is not asking people what they like. Its time to get beneath the surface and find insights to why someone liked or hated something.
Gauging the users' response isn't a straightforward process. In-depth interviews with 5-6 people often lead to a mountain of observations, all of which need to be processed and converted into actionable insights, which further create future design directions. These future directions are the whole reason why such painstaking studies are carried out, which is why we recommend going through an elaborate process of synthesis, even if the insights seem to be jumping right at you from the interviews.
Be patient. Defer judgement. Follow the process. Here it is —
Step 1: Organize Information
The lack of structure and the sheer volume of data can make it easy to lose sight of critical insights.
To tackle this, we individually jot down our observations from user studies onto sticky notes. As we put up these notes onto the wall, we free up our mind to move from storing information to processing it. But before we analyze anything, to keep our biases at bay, we remove all insights and assumptions so that we’re only left with objective observations.
With this in place, we are ready to begin identifying patterns.
Step 2: Find Patterns
Now that we have clearly laid out our data, the next step is to extract the essence of it. To do this, we begin by clustering similar information together. While doing so, we regroup our sticky notes as we notice patterns from our observations.
Lastly, we tightly articulate these patterns into insights by asking ourselves how it relates to our project.
For example, if there is a pattern of users not understanding labels, the overarching insight is that the vocabulary used doesn’t fit their mental model.
Translating patterns into insights
While translating patterns to insights, we articulate the underlying reason for the behavior / need uncovered during research. We do this by asking 'Why?' at every step. Doing this helps the data take a leap from information to insight. For example, a behavioral pattern might be worded as -
Patients do not return to hospitals for follow up care. This worsens their health condition.
An insight derived from this pattern is:
Patients find it inconvenient and expensive to travel long distances to access care. These structural barriers act as hindrances in improving the patient's health.
Actionable insights help the team in making short-term and long-term decisions for the product.
Step 3: Surface Opportunities
A tight research study is marked by clear opportunities for future work. To do this, we mark our insights as critical and frequent. With our most important insights, we write How Might We questions that set the direction for our future design work.
Finally, we look back at our research question and ask ourselves the following questions:
1️⃣Did we prove or disprove our hypothesis?
2️⃣Is there a pattern that suggests new design considerations?
3️⃣What gaps in our knowledge have we uncovered that we need to research later?
By reflecting on these questions, we not only understand the outcome of our study, but we also identify future design and research opportunities.
Insights about a business or product are gathered over multiple research studies with our users. As a result, these insights are often fragmented into their corresponding reports and artifacts. Backtracking each of these insights takes effort. To avoid teams spending hours in navigating through this information, we maintain what we call as meta-insights. Meta-insights are learnings distilled from various research outputs into a single dynamic document. Here is where our qualitative and quantitative data go together, building a single bigger picture. We link each insight to a reference point for teams to be able to quickly switch between macro and micro level details. Each meta-insight is articulated in a way that inspires action, giving the team a clear target to aim for.
Step 4: Share Takeaways
To share our research findings with the larger team, we write a tight report with our most promising insights and opportunities.
A good report needs to be specific — for example, “it is hard to purchase shoes” is unhelpful but “it is hard to filter out the right shoe” sets a foundation for future work. In addition, by backing these insights with data, quotes or recordings, we are left with clear and descriptive insights.
Different ways of presenting findings
- Summarizing findings into a report that substantiates 'what worked well', 'what worked reasonably well' and 'what didn't work well' along with Recommendations and Next Steps
- A findings deck containing annotations on specific screens helps mark interesting behaviors/ Quotes/ Design recommendations. Post-its next to the slide to facilitate emerging ideas, research Qs and next steps for different stakeholders (business, design and product)
- Quantitative visualization tools can at times help organize information at a high level and make it easy for stakeholders to get an overview (Ex: bars to plot users on a spectrum, axis, matrices) https://www.figma.com/file/UEkmvCZdBEPHAC1qO0tPgz/🔁-Iteration-10?node-id=1341%3A
- We find telling a compelling user story an effective way of communicating learnings from the field. We tell a story using thoughtful tools such as story board, maps, and diagrams.