Follow Datanami:
February 16, 2016

What Design Thinking Means for Data and Analytics

Tom Pohlmann

(Sergey Nivens/

From industry giants like IBM to startups like Airbnb, design thinking undoubtedly is top of mind for business teams. The traditional “If you build it, they will come,” mentality has been usurped by techniques like customer journey mapping and empathy-driven prototyping. The user, not the product, leads.

But while design thinking isn’t exactly new and has many use cases outside of product development, many companies are unsure how to implement the approach in a way that improves their business – especially in still nascent areas like data analytics and decision sciences.

At its core, design thinking puts users and their needs as the starting point of developing any new product or solution. The first step is to ask: for whom are we designing and what is the problem they are experiencing? The second: to what end are we modeling the design – to boost consumption and engagement, improve performance, or to achieve scale?

If you think about it, these same questions need to be asked at the outset of any analytics effort. Professionals outside of the traditional product design domain, including data or decision scientists, have the potential to become great designers, they just lack the basic instruction to unlock this sleeping giant within them and apply it in a data-intensive context.

With that in mind, here are the five simple steps that are key to infusing analytics with a designer mindset.

1. Create a Design Framework That Allows You to Fail Fast

Companies may be quick to dive into problem-solving without establishing a proper framework, but it’s a mistake. This can lead to productivity losses and countless hours and money spent on products that should never have been designed in the first place.Design_Thinking_1

A simple framework like the illustration from Stanford University’s to the right sets parameters that can be applied no matter what type of product, idea, or design you have in mind. This framework allows you to stay on track and focus on the problem and customer for whom you’re designing. Just as importantly, a design framework allows you to fail fast and fail cheaply before wasting resources on trying to fit a square box into a round hole.

2. Empathize With Your Customer to Impart Emotion Into Your Product

Design thinking flips the traditional model on its head by identifying the pain point and building a product that fulfills a need or solves a problem. As problem solvers, think of empathy as a muscle that needs to be developed through testing new ideas and hypotheses.

The left-most column of the design framework above is solely dedicated to finding out the real issues and exploring the users’ emotional responses to those issues. In this stage of the design process, the company is empathizing with the user and deeply understanding everything about what the customer needs and Design_Thinking_2wants. To get insight into your customers, conduct interviews, surveys, focus groups and observe users in their daily environments. Diligently take notes and review footage. The results may surprise you.

All of these tools gather insights that are then used to define the problem and build the solution. In the end, the product offers an emotional value proposition, such as a sense of peace, productivity, and ease that was designed from the get-go.

3. Focus on Problem-Solving That Allows for Rapid Experimentation

After uncovering insights from your empathy research comes the hard work of defining the problem. The problem should be expressed in a clear, succinct sentence that combines who the user is, what they are trying to do, why they want to do it, what’s stopping them, and how it makes them feel. The problem statement should have a distinct point-of-view and inspire rapid experimentation.

Bank of America’s “Keep the Change” program provides a good example. Initially, when the team set out to help customers increase their savings, they found that the challenge didn’t align with the bank’s programming, but that saving habits are difficult to change if the burden is placed on the customer.

Using design-thinking principles, the bank tested different ideas where saving required little effort or thinking on the part of the customer. Ultimately they came up with a debit card that automatically rounded up each transaction to the nearest dollar and deposited the change directly into a savings account. The transaction was automatic and, since it was just extra change (not dollars), the customer hardly noticed the impact. The bank also benefitted. As a result of the service, Bank of America claims to have won five million new customers, seven million new checking accounts and one million new savings accounts, all while helping customers build up savings totaling $500 million.

4. Employ Methods to Inspire Creative Brainstorming Across Teams

Tap into the cross-functionality between teams when brainstorming solutions to the problem. Remember that design thinking is a team sport. Not long ago, marketing departments and analytics didn’t mix. As problems within businesses have grown increasingly complex and muddy, integrating viewpoints from multiple teams and team members who don’t all think alike is vital to the success of your product. A good way to unlock insights across teams is to use role-playing, where one group plays the user, another the product developer, another the marketer, and beyond.

If you have data scientists at your disposal, use statistical techniques and regression methods that can help surface unforeseen factors and provoke fresh ideas. Machine learning techniques and hybrid models, like decision trees, can map out different variables based on their importance, which can also inform the design thinking process.

5. To Design the Killer solution, Let Nature Be Your Guide.

Adrian Bejan’s well-known book Design in Nature argued that patterns found in nature have the tendency to evolve toward maximizing flow. A river, for example, branches out into many networked tributaries that allow water to travel greater distances. Looking at the image below, you can begin to spot configuration patterns that evolve in such a way to provide easier access to the currents that flow through it. The best design solutions are no different: they address the current need while simultaneously enabling the flow of future opportunities. When you design with natural patterns in mind, you end up delivering a solution that opens up new possibilities (and products) down the road.Design_Thinking_3

Airbnb provides a great example. In 2009, the company was nearly broke and scrambling to figure out why the company wasn’t growing. One afternoon, as the team was poring over search results for New York City listings, they noticed a pattern of grainy, low-res photographs that prevented customers from getting a good sense of the listing. As an experiment, the Airbnb crew traveled to New York, cameras in hand, and took hi-resolution images of the properties for its users. In a week, their revenue doubled. Airbnb then grew a large network of professional photographers who now help Airbnb users take beautiful, high-quality shots of their properties at no cost. As an added perk, the professional photographs act as a seal of approval from Airbnb, which increases trust in renters. The professional photography service unlocked the “flow” of transactions and even grew into its own small business within the larger home-sharing company.

Just as systems in nature must evolve to survive, the challenges facing today’s enterprises grow increasingly complex with every passing day. However, with the right design framework and these five steps in mind, you can help expedite the product development process and awaken the sleeping design giant within you.

What other companies have you seen successfully design solutions? Have you benefited from design thinking? Share your thoughts or experiences below.

About the author: Tom Pohlmann is Head of Values & Strategy at Mu Sigma, a decision sciences and analytics firm helping Fortune 500 companies to make better, data-driven decisions.