Harmonizing Data and Design:
The Power of Synthesis in Unleashing Creativity
This is Article 1 of 3 from the Synthesizing Information class at the Stanford d.school. The teaching team and co-authors are Richard Braden, Tessa Forshaw, and Jake Hale.
When starting a rigorous design project of any kind — from organizational design, human-centered design, or product design — one must include robust research to understand the problem, the stakeholders, and the need. This design research process essentially follows three consistent steps:
- Collect — Gather information through interactions with people, such as interviews, tours, drawings, and observations
- Analyze — Scrutinize that information by reading, organizing, clustering, and selecting the information
- Synthesize — Make sense of the information and use it to inform a big leap in your view of the world
While all three steps are essential, synthesizing is a special kind of alchemy. It requires you to leverage the collective intelligence of the group to make sense of the data you’ve acquired and then transform those insights into actionable knowledge. Intelligence and data alone aren’t enough. Synthesis is a crucial step that helps you uncover insights and better understand what you’re solving. It also informs the creation of a new idea — a hypothesis — that you can then go test.
But synthesis is a tricky step that’s hard to get right. There is a range of common synthesis mistakes that are easy to make and which can be explained by the cognitive sciences:
In each of these cases, you end up with solutions that don’t test well, connect with users, or solve the problem you set out to tackle. It can end up being like playing the lottery. Sure, you could win and create something that people love, but the odds are stacked against you.
If you get it right, on the other hand, you can make bigger leaps and get better outcomes — with less work.
In design-based research, the key is focusing on a few individuals and collecting high quality data from each, instead of collecting mass information from many people. If you can’t solve a problem for one person, you can’t solve it for everyone. (In the design world, we call that solving for an n of 1.)
There are ways to make this synthesis leap easier though, and more fruitful too. By breaking down the process into distinct steps, we can shape our thinking to more consistently get at a better result. Coming up with a plan for synthesis ahead of time informs the data you collect and the analysis you do.
We designed each class of our six week series to go through the collect, analyze, synthesize phases. We introduced different moves for each phase to intentionally separate the steps. Doing it in this way guides students to dedicate intentional time to each step, allowing them to build on their previous thinking from one step to the next. It also ensures that the data collected in the first phase is sufficient for the synthesis moves planned.
Here are two sequences that we used with students to help with the synthesis process.
Sequence 1: Interview → Empathy Map → POV
This is the most common sequence we have seen taught at the d.school and is frequently seen among many design thinking 101 resources.
COLLECT
This sequence starts with collecting your data, through interviewing users and potential users. Begin by developing an ethnographic-style interview guide to ensure that you capture the most relevant information during the interview. The objective is to listen and understand how the person thinks and feels about a topic. When you ask questions, don’t forget to record what the interviewee says, either via a voice recording or ample note taking. Be sure to ask follow-up questions to get deeper into their answers.
Try to create “thick description” notes. This means recording not only what they said, but also what they did. Write down direct observations or quotes without adding meaning. After you’ve completed your interview, and only after, add some sense-making and meaning-making notes. This entails observing the movements and activities of those you interviewed and interpreting or inferring what conclusions can be drawn from these observations.
ANALYZE
After collecting your interviews, analyze your data by visualizing all of your notes. Put them up on a whiteboard (or whatever space works best for you) and then start digging in by using empathy maps. Pick a user, and using four quadrants, note what they say, think, do, and feel. Think of say and do as a record of data points, whereas think and feel are inferences. Every inference must be linked to one or more data points, but you can have more than one inference from a single piece of data or combine data points to create a single inference.
Do this a few times until you’ve created several empathy maps. Laying out the data from your interviews in this way can help you make sense of all the information you’ve gathered, making it easier to draw conclusions and form hypotheses later.
SYNTHESIZE
Then synthesize by creating a point of view framework. The idea is to encapsulate an insight you derived to help guide your thinking, moving you from data to insight to inference to hypothesis. Start by describing the person you spoke to (We met…). Highlight the elements of your interaction that surprised you, perhaps an insight that caused you to change your own thinking (We were surprised to notice…). Note what inferences you drew from this observation. The objective is to pull together a picture of what’s going on inside that person, getting under how they think and feel (We wonder if that means…).
Based on this information, the last step is to create a hypothesis in the form of a problem statement (It would be game-changing if…). The problem statement can be defined as something relatively open-ended that you think would be game-changing for your user. The idea at this point is not to dictate a particular solution, but rather to use this statement as the basis of your design challenge, allowing you to begin brainstorming, prototyping, and testing solutions.
Sequence 2: Journey Mapping → Tension Dimension → POV
Tension dimensions are often challenging for students, so we designed a sequence and scaffolding to make these tools more accessible.
COLLECT
You can start collecting your data in this sequence by using a journey map to understand the user experience. Either co-create this with the person you’ve interviewed or use the information gleaned from your interview and code it. In the process, focus on uncovering pain points and moments of delight.
To bring this to life, think about the illustrated example above of airplane travel. How did it start? What did the user do? Where did they go? What happened next? Make note of every step of the journey and then add in the emotions they felt every step of the way. Include even the smallest of details, because sometimes that’s where the greatest insights hide.
ANALYZE
Once you’ve compiled all your notes, analyze them and start to pick out the tensions present throughout the journey. Start with the high and low points in the journey map. What did they love or hate, enjoy or get annoyed by? As you reflect on these high and low points, what contradictions, qualifications, opposites, discrepancies, inconsistencies, and surprises do you notice about them?
To capture these ideas, think about words like “but,” “yet,” “still,” “however,” and “although” to get you started. If we revisit the travel example, someone might love the payoff of getting to their destination, but consider the process of the journey to be a hassle. It’s about noticing elements that may not be opposites but are held in tension with one another.
What are the stickiest tensions you’ve uncovered? Turn these into a tension dimension. Create a 2x2 on a piece of paper or whiteboard and pick a spectrum for each axis with opposites on either end. For the travel example, the x-axis could be painless → hassle, while the y-axis might be boring → fun. Plot items on the map. Do this a few times until you end up with about three tension dimensions.
A 2x2 is a great way to make sense of qualitative data. When you force a categorization on top of information, it’s easier to work with it and begin to see relationships. It also allows you to see the impact or effect of two independent variables — how do they interact with one another? All of this enables easier and quicker processing of large amounts of qualitative data.
SYNTHESIZE
From there, synthesize by repeating the point of view process. The tension dimension unearths surprises that can help inform that part of the point of view worksheet. What choices has a person made in spite of the tensions present? Like before, the focus is on the final step of the worksheet, which will dictate how you move forward with your work or challenge.
Building Sequences for Synthesis
We’ve seen these sequences work, helping students avoid synthesis traps by providing an example of how to break down their thinking. In each case we collect data with a tool that supports good data collection (such as interviews, thick descriptions, and user journeys) and then analyze the data (through tools such as an empathy map or tension dimension), and then finally use a POV as a template to help synthesize the data, bringing together data, insights, and hypothesis.
Following a process like this guides your thinking. You can also be assured that you haven’t forgotten a step and, perhaps more importantly, you know your conclusions are based on evidence and building knowledge. Showing your thinking also works for the collective intelligence of the design team, giving you all an opportunity to raise questions and align on your thinking.
Approaching the work in this way removes brute force and luck from the equation. Instead, it becomes a repeatable method for reaching your goal. It’s also a hell of a lot easier: your efforts become more directed and effective. But that’s not all. With these steps, you can unearth important evidence and insights. This data, in turn, allows you to create an informed hypothesis that can be tested, proven or disproven. This leads to finding deeper, more informed insights. And the deeper the insight, the bigger the leap of innovation — and the payoff is huge.
Note: for more information about the moves described in the two sequences above and more, check out the d.school’s design thinking bootleg.
