Big Data Design Thinking

Solving problems and developing innovations for Big Data with Design Thinking

Data is the new oil. But while the use of oil was relatively clearly defined, there are much more possibilities in data. It’s an endless story.

Almost all companies are facing the challenge of having tons of data from very heterogeneous sources. But often, they lack a clear vision of what the data could be used for.

What are new business models that can be fueled with the data? Which data products can be defined and sold by the company? The solution is: Design Thinking.

The Solution: Design Thinking

The Design Thinking approach has been developed in Silicon Valley, but is now used by companies all over the world. It’s main feature is that it uses interdisciplinary teams to create innovations and find new solutions to urgent enterprise challenges.

All participants of a Design Thinking process are consistently switching to the user perspective. In Data Design Thinking, this can be both the point of view of external customers or internal customers for data products. The approach uses various methods such as interviews, prototypes, mindmaps and situational analyzes.

Design Thinking and Big Data

At first glance, this sounds like a contradiction. On the one hand, Design Thinking and its orientation to people and their needs. On the other hand, Big Data, i.e. algorithms and large amounts of quantitative data.

But the truth is, both themes almost are a natural fit. Many of the key features of Design Thinking can also be found in Big Data:

  • Customer centered: Just as Design Thinking means entering the world of the users, Big Data sources are used to analyze business challenges from the perspective of the users. Data such as app usage data from Mixpanel or website data from Google Analytics objectively show the most important behavior patterns of the users and enable companies to develop new ideas.
  • Focused on results: The biggest problem in many Big Data projects is the huge gap between findings from the data and the actual transformation into data products that offer a real value. Here, design thinking helps by consistently focusing on problem solving and action orientation.
  • Agile: An important intermediate goal of design thinking is a prototype, which conveys a realistic impression of the new problem solution as early as possible. The guidelines are: “show don’t tell” or in line with Facebook’s development strategy: “Move fast and break things“. Big Data projects use notebook platforms such as Jupyter, Shiny or Zeppelin to quickly develop data prototypes as well as share and discuss them in teams.
  • Interdisciplinary: In design-thinking workshops, various stakeholders in the company develop solutions and innovations together. Hierarchies and department titles should not play any role. Data Science is also interdisciplinary from the beginning. It combines programming skills, mathematical-statistical knowledge with the knowledge of the respective area – sales, supply chain, marketing or controlling.
  • Empathy: A very important goal of Design Thinking is to take the users’ point of view when looking at the business challenges. These can be external users, for example consumers, but also internal users, for example other departments, that have access to the data products or models.
  • Data driven: Design Thinking does not happen in a vacuum, but is based on collected data, e.g. survey data from a workshop with users and customers. The data is used to define assumptions, test hypotheses, and then come up with new questions, which will then be answered in a next interview round. Data Design Thinking obviously also has a strong focus e.g. on customer interviews, but also measurement data, tracking data, market research studies.

The Data Design Thinking process

Our workshops on Data Design Thinking are based on the following agenda:

Draft

What would the perfect data product in this segment look like? What could be achieved, powered, … with this product?

First interview

Interview, ideally with future users. Key question: What needs do users have? What problem do they want to solve? After the first interviews, the roles will be swapped.

Second interview

Deepen the results from the first interview. If some points are unclear, try to elucidate the users’ “pain points”.

Interim results

Summarize the findings so far: What exactly do the users of the data product want to achieve? What experiences have you experienced in the role exchange? What things were not known before?

Formulation of the Point of View

This is about summarizing the interim results in a single sentence: user ____ needs something (data product, model, analysis, dashboard) to fulfill his needs because ______ (fill in insights).

Ideation

Based on the POV, the participants sketch five radical problem solutions that could meet the needs of users.

Feedback

After a brief presentation of the five ideas, the participants get instant feedback from their partners. Then the roles are reversed and the partner presents her/his ideas.

Improved solution

The feedback is used to develop a new solution.

Prototype

The participants build prototypes based on the improved solution, e.g. in Shiny, Jupyter or Zeppelin, but also pen-and-paper dummies work. This step is not about perfection, but about creating something tangible.

User feedback

Evaluation of the feedback to the prototypes: What has succeeded? What is not good yet? What open questions still need to be clarified? What new ideas have arisen in the interaction with the prototype?

Next steps

Which additions or revisions should be tackled next?

Reforumulation of the POV

Based on the results so far: How has the POV changed? How does a new POV look like?

This method should be able to produce useful and exciting results within a short period of time – between one and three hours – that could serve as a starting point for the development of new data products in the company.

Image credits: gdsteam (CC-BY-SA)