Why do Dashboards Fail?

1. Dashboard Goes Stale

This is by far the number one reason.

When a dashboard is launched, most see it as a finished project. In reality, delivery should be viewed as a milestone. A beta version.

I like to say that a dashboard is a product rather than a project. It requires a owner. Someone who will continually care for it, and ensure it is properly working.

A brand new dashboard is hardly ever widely accepted. Not because the design is bad or people don’t like the colors. But it has to do with a number of issues that are often outside of anyone’s control.

The most common is because the data is dynamic. It changes. New sources are introduced. Current ones are changed. New tables added to the ETL process. All of these things contribute the dashboard going obsolete.

Dashboards are built to answer one or a set of business questions. Those questions change overtime. It starts with one item. Perhaps users might find that looking at total sales for the month is easier to view in another page. That’s one space that no one is using, when you have five or six of those, then the dashboard is stale.

Most importantly, a design might look good in theory, but when it goes live, it is not what viewers expected. That is perfectly fine. Best practices developed in academia is not always applicable in the wild (Zhang et al., 2023). That’s why a dashboard must be continuously lapidated and polished to continue to answer the demands of the viewers.

Imagine with iphone was still in version 1.

2. The Dashboard Lacks Value to End Users

When developing a dashboard, I am usually working with a representative. They bring their own needs to the table and explain what others in their department do. That is not enough.

Everyone involved in the design will often bring business questions they need answers. When delivered, the graphics provided to answer these questions might not suffice the needs to the department. Consequently, the information is unhelpful to the end uses and fails to add value.

Another issue, which I see it often, is the team leads or department chiefs chimming in saying that everyone should look at a specific metric. While the metric could be important and used by everyone, the dashboard might not be the place they find that number (Rossi et al., 2023)

3. Dashboard Relies on Innacurate or Biased Data

Once, I was working in this dashboard that was displaying wrong metrics. Every single element in the page had something wrong with it. Testing the pipeline, I saw nothing wrong.

When the users began to realize that they had closed a lot more sales than it was showing, they stopped using the dashboard. It was all but dead.

Later, I found out the date field accepted multiple formats. The pipeline didn’t account for anything other than dd/mm/yyyy. So, when mmm/dd/yy would come in, it didn’t count for that specific month.

Even after the dashboard was fixed, it took time to get the users to trust it again. Dashboards displaying unreliable data can have long consequences.

4. The Dashboard Results in Unintended Consequences

Some dashboards are designed to provide performance information and facilitate peer comparisons. This methodology, while great in certain settings, it could back fire and have unintended consequences.

In some cases, the user may be dissatisfied with the data presented by the dashboard. Such dissatisfaction may stem from legitimate concerns about the data or from misconceptions about its meaning or interpretation.

Once, a customer called me to complain about a dashboard being wrong. It displayed four open desks in his department when he had six. After some digging, the data was correct. The problem is that he had never off boarded two people properly causing them to stay active in the roster.

Sometimes, these consequences are outside of your zone of impact, but they still have to be dealt with.

5. Dashboard Difficult to Sustain

Again, dashboards are living artifacts. They should be continuously worked on to ensure their relevance.

In some cases, sustainment can be difficult. If the dashboard is not fully integrated with the end users workflow, it will most certaintly fail. Other sustainment factors include accessibility, complexity of the visualization, or if it requires significant expertise or resources to maintain (Rossi et al., 2023).

Another common issue is the price of licenses. Some platforms are extremely expensive and maintaining could be costly.

Finally, the one common issue that I see under sustainement is data acquisition. If the process to acquire the data is complex and require regular security updates, it is less likely to be alive for long.

Conclusion

Coming to terms that your hardwork could be heading to the grave yard is a pill hard to swallow. Trust me. It has happened to me many times.

There is only one way to salvage a dashboard. It is to constinuosly update and polish it. Verify the data sources, ensure the outputs are correct, and it is still answering questions.

At the same time, knowing when a dashboard is no longer sustainable is critically important. Maintaining something that doesn’t work or provides no value is not ideal by any means.