The Data Visualization Dashboard Everyone Wants, but No One Needs
A user-centric approach to designing analytics
What’s the first thing that comes to mind when I say enterprise data analytics and visualization? For many, including myself, it’s dashboards and BI tools. Tools like Power BI and Tableau have become the standard for enterprise analytics and visualizations.
Admittedly, I’m as guilty as anyone else. I have spent hours and hours crafting dashboards, thinking that clients would love them and that they would drive essential business decisions. But the question that lingers in the back of my mind is always, “how frequently do clients look at these dashboards, and do they find them valuable?”
Suppose company ABC hires us to design an analytics package to help their executives understand their current workforce. The go-to approach is to create a dashboard within a standard BI tool and have executives visit it. The dashboard may contain descriptive analytics like:
- Current employees
- Requisitions
- Attrition / terminations
This seems like a reasonable approach and is undoubtedly better than not having access to analytics, but it can get out of hand when the pendulum swings too far. I’ve seen a department with over fifty dashboards, some high level, some detailed, some overlapping, some unique. It’s mind-boggling how anyone could keep track of the existence of this information, let alone make decisions from it.
We need to pause and shift our focus from a data-centric approach to a user-centric one. By understanding how our executives think and make decisions about their workforce, we can better design analytics that genuinely aids them in their decision-making process and eliminate anything that doesn’t help them. Without shifting to a user-centric approach, it’s difficult or impossible to know if our solutions add value or are just another report that never gets used.
The only way to learn from our end-users is to spend time with them. It’s tempting to utilize research from a past project, academic paper, etc., but every company and group of individuals are different and have different needs. Direct research is the best way to understand the unique needs of our end-users. Our approach can vary: The best approach is shadowing. The minimum approach would be weekly reoccurring meetings. A single meeting is not adequate because it will not allow us to understand their day-to-day activities, strategies thought processes or ask follow-up questions.
Here are a few questions that I might ask an executive during an initial conversation.
- Can you describe your role in the company?
- Can you walk me through a typical day?
- Can you describe what workforce management means to you?
- How do you currently keep track of your workforce?
- How often do you talk about or have to make decisions on your workforce?
- Are decisions on the workforce made by yourself, with your staff, or a combination?
- Are there any aspects of workforce management that are more important than others?
- How do you know when a change is required?
- Are there any surprises in workforce management?
- What tool(s) do you currently use.
Anyone who has done research realizes that there is hours worth of conversation in the above questions. This is why having follow-up or reoccurring meetings is critical. The above questions are somewhat high-level; they do not dive into the specifics of workforce management. To be successful, we need to have a rich understanding of our end-users. Each time we meet with them, we solidify our knowledge and refine our approach to meet their needs.
All of this might seem unnecessary, and the temptation will be to build an analytics solution with existing data using existing tools, but consider, what if we learn that most of our end-users read their email first thing in the morning, perhaps a morning summary email that highlights the most critical metrics from our research or potential problem areas would be a more effective solution than a dashboard. Or maybe a text message when something goes wrong or is trending in a particular direction. By designing our solutions to fit the way end-users think and work, rather than forcing them to modify their processes to accommodate our solution, the result will be more successful.
Dashboards have a purpose, but they should not be the de facto approach. By taking time to understand our end-users needs and decision-making processes, we will design more effective solutions that deliver information into the right person’s hands at the right time using the most suitable methods.
