My worst data visualization work has all been dashboards, and I recently realized why.

Part of learning Data Viz is experimenting with different types of visualizations to present information, and dashboards are something I’ve been looking into as I’ve started working with more complex datasets.

It’s a standard visual format that a lot of people have come to expect, but there are several catastrophic mistakes you can fall into when creating one.

To elaborate on this, one needs to look no further than Jared Spool.

Photo by Einar Storsul on Unsplash

Dashboards can easily become a data graveyard

“Dashboards are where data goes to die.” — Michael Solomon, Product Strategy Director

Jared Spool, the founder of User Interface Engineering, tweeted out something last year that sparked a fair amount of discussion about dashboards.

His argument was simple: Dashboards report on the current status. Users don’t act on status: they act on a change in status.

Dashboards are passive when the user needs something active. Without a dedicated user that takes that information, understands it at a glance, and takes action with it, dashboards are more likely to be pretty displays of information.

Does this mean that you should never use dashboards? I haven’t made up my mind just yet. But I’ve come to realize exactly who dashboards are for by asking questions.

Are your audience members Subject Matter Experts?

Photo by Claude Gabriel on Unsplash

One of the most ubiquitous forms of dashboards is a physical one, such as a car or airplane dashboard. But with this type of dashboard, we can see how easily information is extraneous for a lot of users.

For a lot of people, we don’t need this level of detail when we’re driving: a simple speedometer and gas gauge would usually be enough for us for most of our daily lives.

In fact, sometimes this extra information is harmful: I’ve gotten a little stressed when a warning light comes on, even if it turns out to a faulty sensor. Your audience often doesn’t need these extra visuals: it may just be better to focus on one or two charts.

But there are also those that would make expert use of all of the provided information. Racecar drivers or car enthusiasts, who want to know the RPMs. Or mechanics, who can quickly diagnose what a warning light might be. These are subject matter experts that this dashboard is meant for.

So think about the type of audience you’re presenting to: are they subject matter experts? Because then you can ask yourself the next question: does the dashboard provide them the ability to do a comparison?

How easy is it to monitor conditions?

https://www.aitrends.com/selfdrivingcars/speed-limits-and-ai-self-driving-cars/

One of the most common actions we’re likely to take when interacting with a car dashboard is to compare our speedometer with the speed on the road: in that split second, we’re doing an evaluation to compare how our current actions match with where they should be.

This is the type of action that dashboards are intended for. To properly use one, your audience needs to be able to tell the difference between current status and a benchmark just at a glance. This benchmark may be common knowledge to subject matter experts (“All highways are usually 60 mph”), but oftentimes it’s necessary to spell out how the status has changed.

The dashboard on the right can probably only be used by one or two people, simply because they don’t have any context provided: the audience would have to have a specific context already in their mind to quickly evaluate these stats(for example, that our goal was to reach 4,000 patients this month).

The dashboard on the left is slightly better as it shows a number of changes from a baseline, but the audience still has to evaluate whether or not this is reaching intended goals.

Is the information compelling enough to drive action or exploration?

The most important question to ask though, and the one where most dashboards fall flat, is whether or not the dashboard you show drives action.

In essence, dashboards are displaying information (and little else) for your audience to interpret by themselves and take action on. This requires either an extraordinary dataset or an extraordinary audience.

Does this mean anything to you at a glance?

If a data point was completely failing, then we might notice it. Or if our audience were a Google Analytics expert, they might notice something as well.

But oftentimes, this dashboard view is either too complex or not engaging for the typical person.

In that case, what can you do with these multiple visualizations?

Create an infographic.

Why an infographic is usually better than a dashboard

Infographics are usually better for your visualization needs for one main reason: you’re often telling a narrative.

Rather than let the information guide the conversation, which results in your audience having to decipher what it means, you can arrange the information in a way to tell a story.

An ‘info-poster’ of an infographic

This additional structure helps users make sense of this information and turn it into a narrative that is easier to understand and remember.

Does that mean that dashboards should never be used? No: I’m recommending that they not be used in isolation.

I can take a look at this infographic by itself and have a good idea of what it’s trying to say.

With dashboards, I would probably need to incorporate them into a larger presentation or walk the audience through a dynamic presentation. It seems like a complex solution for a complex issue, which is something that has its’ uses but not really for general visualization.

It can be tempting to think about using dashboards, as they are one of the most powerful tools for situations where you need information at a glance.

But it’s usually not the right answer for a general visualization: too much of it is dependent on both the dataset, the type of audience you’re presenting to, and how much time you’re willing to invest.

After all, sometimes the simpler visualization is the better one.

Kai Wong is a UX Designer, Author, and Data Visualization advocate. His latest book, Data Persuasion, talks about learning Data Visualization from a Designer’s perspective and how UX can benefit Data Visualization.

The UX Collective donates US$1 for each article published on our platform. This story contributed to World-Class Designer School: a college-level, tuition-free design school focused on preparing young and talented African designers for the local and international digital product market. Build the design community you believe in.