Understanding Data Driven Design Thinking

TLDR Summary

Design Thinking and Analytics are both used to solve problems. While Design Thinking uses a qualitative approach and Analytics relies on numbers and data, the process they follow are similar. Both can be merged to deliver synergy. The one thing to keep in mind is that both are unique tools for solving problems and cannot be replaced, but should be used in conjunction for maximum ROI.

Two Tools

Design thinking and analytics are two unique and powerful tools in any business context. Design Thinking builds on qualitative aspects and gut feel to make any experience as effective as possible while analytics works on hard facts and data to lend credibility to the effectiveness.

Let's look at how both Design Thinking and Analytics can be applied together for symbiotic success.

What is Design Thinking?

At the grass roots level, Design Thinking is an approach that can be applied to solve any problem in any field - from user experience on a website to the adoption of a mobile app to planning a marketing strategy. The key tenet of Design Thinking is simple: Put the user first.

When design thinking is applied to a problem, it centers on the users. Unlike traditional approaches, the solution is custom-made to suit the user - not the owner of the system. This approach generally increases the effectiveness of the solution by baking in the needs of the user into the solution early on in the process.

The Design Thinking Process

Design Thinking generally follows these steps:

1. Understanding the user: Grok the user and understand their problems, motivations and psychology. An solution that is crafted will need to factor in these data points into the solution. Design Thinkers usually collect this information by talking to the user and observing them (primary research) & looking at data (secondary research). The general outcome of this step is the user persona(s).
2. Define the problem statement: All the research done needs to be distilled into a well defined problem statement. Problem solvers need to pay particular attention to this step and ensure that the problem statement is as accurately defined as possible. Sakichi Toyoda's 5 Whys technique or an Ishikawa diagram are useful tools for this step.
3. Brainstorm: There can be many ways to solve a particular problem and as many of them need to be captured as possible. As with any creative pursuit, any solution, irrespective of how ludicrous it sounds, needs to be captured as each idea can open a new perspective on the solution.
4. Triage & Prototype: Work through the ideas and select the ones (4 to 5 ideas is a good starting point) you can prototype. Factor in opportunity costs, available resources and skills as well. Prototyping can be done on any medium that allows for rapid iterations. You could use mockups and wireframes for digital products or thermocol and 3D printing for physical products. Again, you're only limited by your creativity and skill set.
5. Test & Iterate: Share the prototypes with your users and collect feedback. This is where the final solution will take form as you implement the user feedback and each iteration brings you closer to the final solution. Repeat until the user is satisfied.
6. Develop & Release: The final product is developed and released into the market. In fact, even the release of the product can be mapped by the design thinking process.

What is the Analytics process?

There are many business problems that can be solved with an analytics based approach. Analytics can be applied to any domain and sector where data can be collected - and this is the strength of the analytics approach. Instead of relying on gut feel and intuition, most decisions can be backed by solid data that demonstrates the validity & effectiveness of the solution.

Whether it be predictive, prescriptive or descriptive analytics, in a business context, any analytics project follows a path similar to the Design Thinking process. 

Steps in an Analytics Project

1. Defining The Problem: The entire effort of an analytics problem goes towards finding a solution to the problem. The problem could be as open ended as "How can we generate more sales?" to something as specific as "What can we do to reduce the customer wait time in our support calls?" The problem is defined by the business functions and is translated into a problem that analytics can solve.
2. Data Collection & Preprocessing: Depending on the problem statement, data needs to be collected. In most cases, data is housed in silos in different departments - or even different companies. These data sets need to be integrated and the data collated for preparation. Once collated, the data needs to be cleaned and prepared for analysis. In most analytics projects, over 70% of the time is spent on preparing the data for analysis.
3. Building Models: To build a model, the analyst defines the nature of the problem: Is it a forecasting, classification or an optimization problem. Accordingly, the analyst chooses relevant statistical or machine learning techniques, suitable algorithms and tools to solve the problem. Then, several suitable models are coded and developed. 
4. Testing and Analysis: The analytics model is a theoretical concept and it needs to be tested. Once live, the performance of the model is closely observed to identify different ways the solution can be optimized. These inputs are factored into the next iteration of the model. This process is repeated until a satisfactory solution is found. Sometimes, models may be rejected and the process is repeated till the model works in business scenarios.
5. Business Insights & Presentation: The results from the model are analysed to derive insights. Analysts look for metrics that are significant to the business and translate them into insights that can be applied to business. These insights are woven together with visualization and storytelling techniques to make a compelling case for stakeholders and presented, ending the lifecycle of an analytics project.

Design Thinking & the Analytics Approach: Similarities

When you compare these two approaches to solving a problem, some similarities are apparent:

• Both approaches rely on research for a clear definition of the problem.
• Both approaches start off with multiple proposed solutions (prototyping and models)
• Both approaches arrive at the final solution by iteration

Design Thinking & the Analytics Approach: Differences

Though both approaches are used to solve a problem, they are unique in the following ways:

• Design Thinking relies heavily on qualitative inputs and user feedback while analytics is data driven and is usually applied to a system as a whole.
• Design thinking can be applied to any process and starts from a place of empathy while the analytics approach is completely dependent on the availability of data.
• Design Thinking is generally applied at a holistic level and led by strategy while analytics is more surgical and focused in its application.

The Hybrid: Data Driven Design Thinking

Traditionally, Design Thinking and Analytics have not been used in conjunction. However, both can complement each other. In fact, in my opinion, Analytics can improve the effectiveness of the Design Thinking process by introducing rigour and hard facts into the mix. This hybrid is the Data Driven Design Thinking process.

Data analytics can be used in critical junctures of the design thinking process. In the research & user understanding phase, analytics can tease out hidden insights. Analytics can also be used to process inputs from a larger sample for a more comprehensive and representative user persona.

Analytics can be useful in defining the problem statement as well. To achieve the most granular definition of the problem, we can use Sakichi Toyoda's 5 Whys technique to drill down to the root cause. However, the answer each to 'Why?' may not be apparent and would need a deeper study. Analytics can help here by identifying hidden relationships in the data.

Next, analytics is the core of the test and iterate approach. It can successfully bolster the qualitative nature of the Design Thinking process. It adds rigour and develops solutions based on facts and data. Analytics can, in most cases, objectively determine the success of a proposed solution and recommend improvements to unsuccessful ideas.

Finally, analytics, along with data visualization and storytelling techniques can help problem solvers communicate in a way that would appeal to all stakeholders and promote acceptance of the solution.

Caveat

One thing to remember is that Design Thinking and analytics are approaches. They are tools to apply while unpacking a problem and solving it. They complement each other strongly, but do not replace each other. They function differently and serve different purposes.

Summary

In summary, this hybrid, Data Driven Design Thinking, can be considered a spiritual successor to Design Thinking. Analysts and Designers can benefit by applying principles from the other field to round out their approach and deliver a more comprehensive solution as demonstrated by companies such as Slack, Google, Facebook, LinkedIn, Ideo and a number of other members of the new economy.

Author's Note: Please contribute your opinions in the comments to further the exploration of this topic. I'll respond to each comment. Thank you.

Photo Credits

• Photo by William Bout on Unsplash
• Photo by Matt Artz on Unsplash
• Photo by Carlos Muza on Unsplash
• Photo by PICSELI on Unsplash