In the first two parts of this series, we explored how even the best-engineered data platforms can be built for failure and how the data awesome framework can leverage design thinking to create a people-centric data solution. Here, we’re going to share some of the lessons we’ve learned in putting this into practice. (Read Part One here; and Part Two here)
The Data Awesome mindset can be summarized in one simple sentence: Make everyone data awesome. This mantra brings a human-centred approach to data and a customer obsession in our data products. We measure success by the extent to which we can change how people feel about data: from data frustration to data empowerment.
Data frustration is the term we coined to explain a situation when data experience is below the users’ expectations. By this we mean when users perceive data to be a waste of time and energy, feel overwhelmed by the data or fall back to their comfort zone (e.g. spreadsheets) if their expectations haven’t been met. Data empowerment, on the other hand, is when the data experience exceeds the users’ expectations and enables them to complete the outcome they set out to achieve.
We started with just one project, one dashboard for staffing leads to enable better supply planning. After a few weeks usage quadrupled, feedback went from “I can’t make head or tail of it” to “I can’t work without it”, ‘I am ready to drop my spreadsheet’. The new dashboard was tailored to the needs of its audience. It empowered them to be better at their jobs. They loved it and they not only adopted it but they became champions for our data program. That dashboard raised the bar for all the others in terms of usability, usefulness and value to the business. Stakeholders and tech teams started to use that as a reference.
In another instance, we didn't have enough data available to prototype a dashboard so instead we paired with the Global Head of Supply to prototype the slides she was using for her monthly presernaton to our Global Leadership Group. Rapid prototyping delivered immediate value to the business and played a key role in converting our stakeholders into partners and getting dedicated time for this data project. By doing our designing on a slidedeck, away from any data availability constraints, we created a shift in the team’s mindset. It is not about the data available, it is about what data helps stakeholders to get better at their jobs. This exercise was instrumental in shaping priorities for our data products. Rapid prototyping applied to data saves time, money and avoids over-engineering. Benefits can be measured as a reduced cycle time to generate answers to business questions.
Too often, when businesses try to become data-led, they think it’s about giving their staff data dashboards instead of thinking about what data they need to do their job. One common pitfall is starting from the data and not from the people: this means that they share data, not insights. We need to understand what information staff needs to get their job done and what business outcome they are after. Only then we can shape the data into meaningful insights. Insights are made of a question, an answer and enough context to make the answer meaningful.
Data visualization is how we communicate the answer to shape the insights, a good data visualization is the graphic representation of the answer that is most effective at sharing a message. Anyway, if your data is not answering the questions that users need and want to do their job, your data is useless, even if you are using the best data visualization.
Are you familiar with these situations? Business stakeholders don’t trust the data. Business stakeholders don’t engage with analytics. The Business doesn’t see advancement toward becoming a data-enabled organisation despite the technical data effort and the investments. Engineering team and business stakeholders are disconnected.
The Data Awesome Framework bridges the gap between the technologists and the business by leading them to a path of collaboration through the understanding of business needs and the co-design of the insights. It provides them with things to do together and by themselves to produce early value. Engineering teams develop better empathy and understanding of customers’ needs and use it to prioritize the technical challenges that need to be solved first. Engineering teams and business stakeholders partner to co-design and produce analytics. A human-centered approach creates short loops of feedback to foster a partnership between tech and business and a shared sense of ownership. Better engagement of business stakeholders results in improved efficiency and continuous improvement of the business and operations.
Making everyone data awesome doesn’t mean that everyone needs to become a data scientist or a business intelligence expert. Being aware of our team and business stakeholders data fluency can make a huge difference. By sitting with some of our business stakeholders we learnt that they were using interactive dashboards as printed reports, missing out on the potential of filter, crunch and drill. We created shortcut filters directly on the dashboard and we made it obvious when widgets were clickable. Engineering teams were investing a lot of time to learn our Business Intelligence tool to design better dashboards for our users. By working with them to create simple templates and drag & drop widgets to cover the most common scenarios we were able to reduce the time to deliver insights to customers from months to weeks and we registered an 87% increase in the user engagement of the engineering teams while creating new dashboards.
Key Contributors: Anumeha Verma, Adityan Ramkumar S B, Ying Fu, Ishan Borah.
Special thanks to the team that built the first Data Awesome dashboard: Shweta Nayak, Tiara Xiao, Pei Ren Tan, Chunjing Liao, Shiqi Yang, Wei Cheng, Wen Yan, Xiaoling Bian, Wen Hu, Susie Cao.
And finally all the Thoughtworks Data Awesome Ambassadors: Advithiya S, Almas Munna, Bhaskar Rao, Carol Vieira, Chen Wang, Chenxi Li, Debashis Chandra, Deepika M, Deepthi K, Fabiola Vieira, Felipe Silva, Gayathri Pandian, Jie Liu, Jitendra Bachhav, Jorge Moraes, Krotoka Singh, Qu Huang, R Mathurbhashini R, Rahul Suman, Rathinakumar Ponnusamy, Rijilash Vijayan, Saraswathy Renuga Sankarasubramanian, Si Wan, Sivakumar S, Subramanian CV, Yilun Liu
Disclaimer: The statements and opinions expressed in this article are those of the author(s) and do not necessarily reflect the positions of Thoughtworks.