This article is written by Khushali Sandhi. Khushali is a Senior User Experience Designer at Amazon with over 13 years of professional experience, specializing in creating impactful UX/UI solutions for complex cloud-based database systems. After beginning her career in front-end development in India, her passion for understanding and solving customer problems led her to transition into UX/UI design, culminating in her current role at AWS.

So, you've decided to design a dashboard? That's great! I will provide some suggestions that could help with the process, drawing from my experience in designing dashboards. Whether you're designing a dashboard from scratch or redesigning an existing one, it can alter your approach, but there are general guidelines to keep in mind. Let's dive in.

For this article, let's assume we're designing a dashboard for a sales manager at a cake bakery shop where customers use an online ordering app to place orders.

Understanding the Need

First, we need to determine if a dashboard is necessary. How do we know it's the right solution? Was it proposed by a Product Manager (PM)? If so, why? Are there alternative solutions? If a dashboard exists, analyzing page visits, engagement, and usage statistics can reveal its usefulness to customers. This stage is about understanding customer problems. Sometimes, the request to design a dashboard can lead to premature solutions. 

Instead, take a step back to gather data on current customer pain points, such as difficulties in monitoring resources, identifying resources needing assistance, or spending too much time on troubleshooting. For our cake bakery example, the sales manager might want to track deliveries, group orders for efficient delivery assignments, or identify delayed or incorrect orders. After gathering data, it's time for discovery and exploring high-level concepts.

Discovery (Defining the Goal of the Dashboard)

Understanding customer problems is crucial. Schedule deep-dive workshops with the team, including PMs, engineering leads, customer support, and internal customers, to learn about the data customers seek from a dashboard. FigJam can be a valuable tool for these sessions. 

Remember, everyone has opinions, and feedback may vary based on their team's focus. The goal here is to collect diverse data points without narrowing down on specific problems. Group notes based on common themes, such as monitoring delivery status, customer support needs, and ordering patterns. Conduct a prioritization workshop to rank these themes based on importance, using previous research findings and assumptions to guide you.

These themes help define the dashboard's goals and may lead to discussions on secondary target audiences. It's easy to assume all customers are the target audience, but often, a specific group is primarily responsible for monitoring these statistics. Narrowing down the target audience at this stage helps further refine the dashboard's goal.

Note: Avoid defining metrics at this stage. During workshops, stakeholders might suggest specific metrics, like the average number of chocolate vs. vanilla cakes ordered. These suggestions, while helpful, can distract from the broader picture and should be saved for later.

High-Level Concepts

With the goals defined, it's time for high-level conceptualization. Focus on the dashboard's story, detailing the customer journey through the product. For our bakery example, consider the simple flow from order placement to delivery. Use this journey to build the dashboard's story, identifying key data points like popular cakes, seasonal order patterns, inventory management, delivery efficiency, and customer feedback. This exercise helps pinpoint each touchpoint and uncover new opportunities.

Sketch low-fidelity wireframes to outline this story, using data from previous sessions to refine your narrative. High-level concepts should avoid detailed metrics but lay the groundwork for further discussions.

Defining Metrics

Now, define metrics that will tell your dashboard's story. Incorporate findings from customer research and workshops and consider inspiration from competitors and unrelated domains. Include a diverse group of stakeholders in this process to ensure a comprehensive set of metrics that effectively communicate your narrative.

Ideation

With a clear understanding of customer pain points, a defined story, and set metrics, begin initial designs focusing on layout, metric aggregation, and dashboard filters. This stage involves validating assumptions, iterating designs based on feedback, and potentially planning phases for the dashboard's launch. Aim for a dashboard that regularly aids users, understanding that not all customers will be fully satisfied.

Customer Feedback

Prepare a research plan for diverse customers, considering new and experienced users from various enterprises and use cases. Understand that the dashboard will evolve, and consider a beta release to gather initial feedback.

Define Success

The goals set earlier also imply success criteria. Measure how the dashboard has improved efficiency or solved problems for its users.

Measure Success

Evaluate success through defined metrics, page visits, dwell time, user actions, drop rates, and customer satisfaction surveys. Combine quantitative and qualitative feedback for a comprehensive understanding of the customer experience, allowing for ongoing refinement.

Designing effective dashboards requires a deep understanding of customer needs, a clear definition of goals, and a creative approach to presenting data. By following these guidelines, designers like you can ensure your dashboards are valuable tools that enhance decision-making and improve user experiences. 

Remember, the journey from understanding the need to measuring success is a collaborative effort that evolves with feedback, making each dashboard a unique solution tailored to its audience.

References:

• R. Kosara, "Visualization Criticism - The Missing Link Between Information Visualization and Art," 2007 11th International Conference Information Visualization (IV '07), Zurich, Switzerland, 2007, pp. 631–636, doi: 10.1109/IV.2007.130.
• Z. J. Beasley, A. Friedman, and P. Rosen, "Through the Looking Glass: Insights Into Visualization Pedagogy Through Sentiment Analysis of Peer Review Text," in IEEE Computer Graphics and Applications, vol. 41, no. 6, pp. 59–70, 1 Nov.-Dec. 2021, doi: 10.1109/MCG.2021.3115387.
• Midway, S. R. (2020). Principles of Effective Data Visualization. Patterns, 1(9), 100141. sciencedirect. https://doi.org/10.1016/j.patter.2020.100141
• Cairo, A. (2012). The Functional Art : An Introduction to Information Graphics and Visualization. Pearson Education, Limited.
• Steele, J., & Iliinsky, N. P. N. (2010). Beautiful visualization : [looking at data through the eyes of experts]. O'reilly.
• Wathan, A., & Schoger, S. (2019). Refactoring UI. Adam Wathan & Steve Schoger.