What is Servian’s Data Visualisation design process?

When was the last time you viewed a Tableau or Qlik dashboard and thought – “Geez! I can really make a decision with that information!” Too often dashboards become a collection of semi-related charts rather than a mechanism by which you can inform a decision.

At Servian, we believe that having a practical, repeatable innovation methodology is essential as it brings structure to your innovation process. Over the years of experience delivering 100s of digital experiences for our customers we’ve developed an approach called ‘Rapid Experimentation and Prototyping (RapidXP)’. This approach helps our clients to create the vision of a digital product without needing to key in a single line of code.

Data analytics and reporting products are digital products. They use and visualise data to support decision-making and help make decisions using complex systems. Surprisingly in the data and analytics space for the business intelligence world, the user first approach is relatively new. The traditional dashboard design process focused on functional skills, like how to create box plots in Tableau, not the complex processes of understanding the user needs and gathering business requirements.

We believe that by adapting our Rapid XP process to BI helps our data analysts construct data stories that resonate, and allows for the effective communication of insights. We still believe that there are some challenges when adapting Rapid XP in the design of data products. The data competence and data literacy of the user is often unknown or too low to be involved in the initial design process to effectively scope requirements. Recognising that users often struggle to clearly articulate dashboard requirements, we reimagined the application of the design thinking process and created Rapid XP for Data Visualisation and Dashboard design framework, tailored for data products.

“Rapid XP focuses on using ideation and stakeholder feedback to product tangible outputs. Fostering innovation helps you fail fast and succeed sooner!”

Ready to take the next step? Let’s look into the state of your existing apps and what changes would be needed for a cloud transition.

Take advantage of our FREE 2-hour consultation

Our Three-step design process

Similar to Rapid XP, we will work through three consistent, engaging and collaborative steps while engaging with end-users, stakeholders, customers and management alike. Bringing together a diverse group of people from different areas provides us with an opportunity to get hands-on with their data problems.

1. Empathise & define

 

The path of gathering reporting insights allows us to anticipate the critical business questions users need to answer to enable their decision making. We look to identify key objectives, audiences, and the priority of their business questions. We have a set of methods to understand and frame the key challenges. This process sets the foundation for an information architecture that supports the dashboard creation.

 

2. Ideation: collaborate, visualise, storytelling

 

The design workshop

 

Drawing is a great way to learn from the end users of the dashboards. A quick sketch, graph or concept diagram is a great way to break the language limitations and keep a physical record of people’s ideas. The design workshop is structured as multiple rounds of time-boxed sessions. Each participant sketches one or more ideas, after which they present, discuss, sort and prioritise their ideas to the solution based on how relevant they are in answering user goals and questions. This is a critical process, it helps us to explore the possibilities, get familiar with the business problems and enable the end users to become familiar with the data. Together we make informed choices about the visualisation design, communicate ideas, break down silos and show participants how to challenge their assumptions. We’ve ultimately found that the best outcomes are achieved by running a workshop per dashboard.

 

Toolkits

 

During the pen and paper sketching session, we hand out a deck of data visualisation cards, showing the users common names and usage of those chart types. This toolkit provides a visual example, not only inspiring users on the visualisation possibilities and the best way to tell their story, but also to help them overcome the intimidating feeling of not being an artist.

3. Design and Test

 

One of the biggest challenges in leading the design process to create data products lies in the designer’s needs for deep insight into the data structure. This means understanding the raw data and the data pipeline to better uncover the technical feasibility and scalability of the design – a pure UX view is not enough. Our data analysts work alongside the designer to provide a more holistic view and help the designer explore the data potential and the technical limitations of the BI tool of choice. Pairing designers with BI developers effectively, just as we would pair designers with front-end engineers, is the key to success.

 

Session 1 – Low Fidelity design

 

Sketched ideas and solutions consolidated into a set of wireframes that are intended to capture the overall information architecture, and visualisation to answer the business problems.

 

Session 2 – Validation

 

Using the same participants as the original design workshop, the validation session is the opportunity for stakeholders to give feedback on the direction of the dashboards being designed.

 

Session 3 – Iteration

 

Using the same participants as the original design workshop, the validation session is the opportunity for stakeholders to give feedback on the direction of the dashboards being designed. It is not realistic that the problem will be solved in one workshop (or one workshop per dashboard), we take time to work with the stakeholders and users to thoroughly understand their challenges.

 

Session 4 – High Fidelity design

 

Similar to a web app product, the designer will take the opportunity to visualise the full breadth of the experience. It is not only a UI design exercise, we also cover how the user interacts with the visualisation, how they drill down to find more details, what level of flexibility can the visualisation tool provide to enable different roles of the user. Given the varied constraints of BI tools, it’s imperative to capture the core functionality of the dashboard up-front whilst leaving room for the BI developer to have a level of creative freedom.

 

Session 5 – Design validation

 

Finally, we validate the designs, in the same format as the first validation session. With high fidelity design of the final products, we encourage the technical feasibility discussion to be included at this stage.

 

Design handover to implementation

 

This is the final phase before moving to development, with the high fidelity design and style guide the developer will be able to start prototyping with data and build intuitive, visually stunning dashboards.

 

Authors: Servian Design Team

Interested in learning how Servian can work through the RapidXP process in your organisation?