The Key Thing Wrong with Your Dashboard

We do Data Visualization Consulting.

We've helped large brands and small product companies make compelling cases with data visualizations.

By now, everyone is familiar with dashboards. You likely have a dashboard for your website analytics, your business operations, and maybe even your personal finances. Dashboards are the quintessential way most people interact with data, and often the best tool to stay on top of key metrics and drive decisions.

But dashboards can be overwhelming, and many people get frustrated for one key reason.

Monitoring the situation

A dashboard is just a specific kind of report designed for easy monitoring of key data points. Good dashboards provide clear, understandable data about things you are interested in; and importantly, they make it fast and easy to consume the data.

The data used to populate dashboards is relatively simple. In the most basic version, its a single data point, like total number of visitors to a website. More complex dashboards will begin to show you how key metrics have changed over time, or the amount of change between one period and the next. You can further improve dashboards by segmenting data; showing the breakdown of visitors by country, for example.

Where dashboards fall short

When using a dashboard, you've probably asked the questions 'But what does that number really mean? Is it good, bad or unimportant?' This is the key problem is dashboards: they lack context.

Data in-of-itself is useless. Context gives data meaning. Without it, we can't make decisions.

When looking at a standard analytics dashboard, it is impossible to tell whether the number of page views, the bounce rate, or channel distribution is good or needs improvement unless you are an expert in website analytics. The numbers themselves are largely useless to drive complex decision making; it can be helpful in making incremental improvements, but can't help you understand if your website is helping you outperform your peers.

Context is everything: moving to the scorecard

So what exactly is context, and how do you get it? In our work, the most useful context for decision making is understanding of the wider world you operate in. Specifically, data about your peers, industry, locations, and outcomes.

Website analytics is much more useful if you know how the numbers compare to the industry average, or organizations of similar size, or in your area. With this information, you can make strategic decisions that dashboards don't inform; is the design of my website driving more sales than the competition? Is my return on capital better or worse than the industry average?

Increasingly, we recommend clients consider another approach to monitoring key metrics: the scorecard. Scorecards are inherently better than dashboards because they do more: they include the individual metrics of dashboards, but provide the relative context that helps you make informed decisions.

What makes a great scorecard?

A great scorecard helps you both monitor how you are doing, as well as explore the context behind the numbers.

Specifically, scorecards need to:

  • Show you how your numbers compare to relevant peers. This can mean national, state, or industry averages.

  • Visually rank and order the most important values you need to pay attention to.

  • Show both real numbers as well as relative variance between comparison data.

  • Support drill-downs and pivot operations to see the data in different ways. You should be able to answer the 'why?' questions as well as the 'what?'.

The best way to identify a great scorecard is easy: can you give it to someone who has little background in the subject matter and have them identify what is important?

Scorecard Challenges

So why don't most organizations use scorecards instead of dashboards? Scorecards require a lot of data. You need to have the basic numbers used in a dashboard, but also need the contextual information, comparisons, etc., that the scorecard requires.

Scorecards also require more complex statistical analysis than dashboards. You need to make sure you are comparing apples to apples, and this often requires complex statistical approaches like zScores, distributions and regressions.

Scorecards also present interesting visualization challenges. Making a large amount of information easily understandable is always hard. Effective scorecards are usually heavily customized to support the specific users and decisions that need to be made.

Evolving from dashboards to scorecards

So how do you evolve your dashboard to a scorecard?

The first step is to gather the data; you'll need to identify and collect the contextual information you need to understand to make your decisions. Often, this means investing in collecting or purchasing data from outside parties who can provide a cross-organization or industry-wide perspective.

Second, you'll want to design the scorecard that makes the most sense to you. Since the value comes from regular use, you need to have visualizations that are easy to understand and use.

Finally, you'll need to consider how you keep the data fresh. Scorecards need to be updated frequently to keep the context accurate, and having a plan in place for regular updates will be key to getting the most from it.

Interested in learning more about how scorecards can improve your decision making? We're here to help; send us an inquiry.