After the previous article about Communication With Data, What Is It?, we think that sharing best practices on how to make dashboards business-relevant would be easier for people to inject data viz techniques into operational work. Let’s get started with the simplest element: text.
There are probably many in the data visualization community who have a visceral response to the notion that we should be including more text in our visualizations.
The reasons I gravitated to data visualizations in the first place is because:
- They are visually appealing
- They don’t require the audience to read
- They don’t require me to write
Texts allow for none of those things. Well, maybe a text can be visually appealing if you have a really great font.
One of the first things you learn in data visualization is to get rid of words. Less text, less lines, less noise — less, less, less. So, why am I encouraging the use of more text? Because, to my chagrin, there are some things that are, and always will be, communicated more effectively with text. Them be fightin’ words, but I stand by it.
To be clear, static text has very little business being in your dashboard. Static text is text that is always the same and does not change based on context like filters. Dynamic text, on the other hand, can change based on filters or on the data itself. Here is an example of a dynamic title that reflects the filters being selected:
Title of the visual reflects the department and location filters being selected
The trick to utilizing text to make your visualizations more effective is to use it sparingly and to make it dynamic.
In every single dashboard that I make, I include at least three dynamic texts. They are used to:
- Inform the user about the underlying data
- Draw attention to filters
- Articulate takeaways
Okay, my intrepid warrior, let’s go through exactly why and how to utilize dynamic text to make your dashboards more effective, efficient, and enjoyable.
1. Inform the user about the underlying data
When creating a dashboard, one consideration (of the many, many considerations) is what does my audience need to know about the underlying data? Some attributes about the data that could be important:
- Source. Is the data coming from an internal or external source? Who is the owner of the data?
- Collection. How was the data collected? When was the data collected and/or updated?
- Manipulation. How was the data manipulated from its original source?
- Calculations. How were the metrics calculated?
At a minimum, on every single dashboard I build, I will create a measure that states when the underlying data was last updated. There are a couple of ways to do this depending on the level of granularity the audience is interested in:
- As a dashboard title. HR Dashboard August 2020.
- As a footnote. This data was last updated on August 15, 2020.
2. Draw attention to filters
One of the most useful aspects of a dashboard is the ability to allow users to filter the data to meet their unique needs. However, this can be dangerous if the end-user doesn’t understand how the data is being filtered.
While you might hope that having the slicer on the page would be enough to inform the end-user when a slicer is selected, it often isn’t. Even though an angel dies every time someone screenshots a visual from a dashboard to put into a PowerPoint, this happens all the time. And when people do screenshot, they often only include the graph itself without the filters in the screenshot.
If we did not have dynamic texts in our title, someone might see this in a PowerPoint and think it is full company without realizing it is being filtered just to the Finance Department.
When this happens, people will either:
1. Freak out and think everyone has left the company
2. Freak out and think there is a mistake in the data
In order to save us all from a moment of panic, let’s make our static Headcount title more dynamic.
Let’s say that you have a slicer on your dashboard that allows your user to select different departments that they might be interested in. You also have a visual that shows headcount. To draw extra attention to the slicers being applied, you can include a dynamic text in the title that updates what it says depending on whether or not a department is currently being selected in the slicer.
Now that we have made our Headcount title dynamic, even if someone blasphemously plops a screenshot into a PowerPoint, the viewer will still know that the data has been filtered to show just the headcount for the Finance Department.
3. Articulate takeaways
While I often feel like the takeaway from a data visualization is screaming at me, I have learned that this is not always the experience of end-users. For those of us who are (wonderfully) saturated with data all day every day, we subconsciously consume data visualizations in a systematic way and know exactly what we are looking for. People who have lives that don’t revolve around data have to spend a lot more conscious effort trying to digest what is being presented to them and what they are supposed to do in response.
Depending on the complexity or novelty of the message you are trying to get across, you can create takeaways that are:
- A visualization title. In the example below, the number of hours (50), whether it says “more” or “less” (more), and case type (Benefits) are dynamic.
- A text box. In the example below, the text box below the bar chart has dynamic text that changes based on how the top case type compared to the previous month. It also gives insight into where the majority of the cases are coming from to give the user additional insight.
As you can see, the text box option requires a much more complicated measure than the visualization title and it also adds noise to your visual. You could potentially add additional visualizations that showed the hours spent on each case type month-over-month and the number of cases coming from each department. You could say that you specifically built this dashboard in such a way that an end user should easily arrive at this answer themselves.
I hate to break it to you, but no matter how intuitive, well-built, and beautiful we feel like our tools are, end-users might still need some hand holding. Putting in the extra work to create more thorough takeaways can really pay off, especially when:
- The dashboard is new
- The end-user is less familiar with the data
This is a great reminder of how important it is to solicit feedback from your end-users and find out more about how your tool is actually being used. Do your users know exactly what they need to do once they see their data each month? Are your users taking advantage of all of the amazing insights your tool has made possible?
Used sparingly and intentionally, dynamic text can be a game-changer for your dashboards. To level-up your dashboards, use dynamic text to:
- Inform about the user about underlying data using a dashboard title or footnotes.
- Draw attention to filters using visualization titles.
- Articulate takeaways using visualization titles or text boxes.