The Discipline of Innovation
“reality is an ongoing accomplishment that emerges from efforts to create order” — Karl Weick
Growing up, I had the fortune to study math alongside design. Paradoxically, I learned the most valuable design lessons from the math books rather than the design ones. In particular, George Polya’s book “How to Solve It” taught me how to approach finding superior design solutions in a disciplined way. I don’t mean to imply that there is a solution which is inherently superior to all others, but rather that the “superior” solution is the one that is most optimized for a specific environment.
Let’s take an example which feels too familiar these days — a virus. Occasionally, a small mistake or mutation in its genetic material will give it a radical advantage to survive and compete with the host’s defense systems. Because the process of natural mutation is based purely on randomness, the scientists are often able to outrun the mutations by following the discipline of scientific discovery to develop vaccines.
The same idea applies to business. If the business is only working on incrementally making “horses faster,” sooner or later the business will be outrun by the radical mutation of a “car.” Businesses need to innovate in a structured and disciplined way to survive, and to hedge themselves from the risk of becoming obsolete.
This idea is captured in Noriaki Kano’s model, which dissects customer satisfaction into several categories, the top left of which is the “unknown needs”. Accordingly to Kano, having a portion of your product portfolio in this quadrant can potentially lead to the highest gross profit margin. The question is how to get there.
There are two parts to that — being able to foresee the change in the environment, and being able to take advantage of that change.
Businesses always have a temptation to blindly follow the lead of the industry analysts who are forecasting future markets. However, there are two problems with this. One is that all your competitors are doing exactly the same thing. If you only follow market trends, you will never get ahead of them. The second is that analysts’ research is top-down and aimed at informing buyers and promoting vendors. Our aim should instead be to discover the latent trends in user experience bottom-up. You can say analysts study the shape of the trend wave, while we study what causes the wave’s shape, which I believe, is ultimately people. The deeper and earlier the insights, the bigger the competitive advantage.
However, design researchers nowadays are largely unprepared to studying the deep and dark waters of people’s implicit desires and needs. Perhaps rooted into the classical rationalism of the Age of Enlightenment, contemporary schools prepare us to design single-variable experiments to study user behavior with the notion that there is an objective reality out there and our only job is to understand it. We can track product usage in different market segments and forecast which one will grow faster. We can run an A/B experiment and learn which version of the product is easier to use. We know the dimensions of the problem space, and we have a bias or hypothesis that we are trying to prove or falsify. We know what questions we want to get answered, and we know how to ask those questions.
But the world of design and design research is often trickier than that. Design researchers looking for innovation opportunities are often in a situation where both the product and the user is unknown, thus they can’t define the boundaries of the problem space to start with, or even develop a hypothesis to test.
Moreover, by its nature, the design introduces previously unknown phenomena into the world causing the emergence of new behaviors and changing our understanding of the world itself.
‘‘The basic idea of sense-making is that reality is an ongoing accomplishment that emerges from efforts to create order and make retrospective sense of what occurs’’ wrote Karl Weick.
Forty years prior Heidegger wrote “Techne […] reveals whatever does not bring itself forth and does not yet lie here before us, whatever can look and turn out now one way and now another.”
This is why we need design research methods that are more phenomenological in nature and help us with the emergence of potential patterns of significance while we attempt to make sense of “wicked” problem spaces.
There is a certain technique I came up with over the years that was helpful to me in “emerging” insights out of fog. I find that this technique is similar to Artificial Intelligence trying to make sense of an image.
I’ll demonstrate the technique using a case study.
Re-framing the prompt
A few years ago I joined a team looking into conversational interfaces. My project was to research use cases of conversational analytics, and there were a number of products on the market that would allow users to query data conversationally. Our original plan was to collect samples of user utterances to inform the design of a product just like these, which would have put us on par with the competition. However, we decided to reframe that into a study of how users have conversations with data to solve business problems. That gave us opportunities to look for superior product ideas and to gain a competitive edge.
Inductive coding
Over the course of the study, we interviewed a number of business analysts and citizen data scientists. We asked participants to pick a recent problem they faced and narrate their process of solving it to us.
Once the interviews were complete, I analyzed their responses verbatim using a technique called “open tagging” or “inductive coding”. This involved reading through the text and tagging each excerpt with a few words describing what that excerpt was about.
After that was done, I would logically organize my tags at the top of the document.
This analysis outlined the main themes, or semantic dimensions of each interview for me. The next step would be to compare the tags (dimensions) between the interviews. In the process of synthesizing these dimensions, the patterns emerged.
Design thinking
It became very clear that both business analysts and citizen data scientists follow repeatable processes to solve problems with data. However, business analysts use a cruder process and display more flaws in their statistical reasoning than their citizen data scientist colleagues.
With that, we had a clear vision of where the product opportunity lies: we would “package” the data science expertise, in the shape of conversational guides so that everyone can have access to the same data analysis sophistication and precision.
And so you have it. To bring it back home to George Polya, mathematics, and problem solving, here is a quote from Alon Amit, a mathematician and “a product guy”:
“When you are looking for a solution, it is unpredictable. When you find it, it is unavoidable”
To summarize:
- Businesses need to be making investments into their design innovation pipelines or they’ll be eaten up by their competitors.
- Design researchers shall not be afraid of the wicked research problems.
- There are modern approaches that allow modeling wicked problem spaces in a constructive way.
Credits: thank you Corey Keyser for your review and suggestions; thank you Ana Manrique for lending me the “Design for Dasein” book, reading it led me to discovering even more exciting things to read; and thank you ChatGPT for verifying my sources and editing my draft for me.
Julia Blyumen is a designer at IBM Data and AI based in Silicon Valley, California. The above article is personal and does not necessarily represent IBM’s positions, strategies or opinions.