Strong data storytelling is like a decoder ring that unlocks the insights within your datasets and propels your teams onward, to the decision-making and action-taking that takes your work to the next level. It’s the key to becoming a truly data-informed organization.
The default charts you can produce with your go-to software, like Excel or Google Sheets or even Tableau, will never get you to a data story. Because, even if enhanced with AI, they don’t know your story. You do.
That said, you can reshape those default charts so that the story you see is clear and direct for your team, your audience, and your stakeholders. It’s easier than you might think.
The defaults don't do you justice
Let’s say you're evaluating the effectiveness of your nonprofit's email strategy. You’ve run three types of email campaigns this year: Advocacy, Fundraising, and Newsletter. Now you want to show how they performed over the last four quarters.
A default chart of that campaign data would look like this:
This column chart might not seem so bad at first glance. But try to pull out a story. It’s a challenge that requires some insight, some data literacy, and some mental effort - all of which will clog up your team meeting and slow down your progress.
The reasons this default chart drags you down:
- Vague title: “Click Through Rates by Email Campaign Type” states what the data is about. But it doesn’t tell you the story in the dataset. What a wasted opportunity!
- Overly colorful: A different bold color on each email campaign makes a viewer’s eyes bounce all over the chart, never landing on the evidence that conveys the story.
- Cluttered labeling: The repeated (%) along the x-axis is redundant of the “percent” y-axis label. Additionally, the data labels on every column are redundant of the y-axis altogether! All that excess creates more noise for your viewer to dig through.
- Hard-to-follow comparisons: The chart type itself makes for some needless strain. In order to see how, for example, Newsletter performed over the 4 quarters, a viewer’s brain has to try to un-see the blue and orange columns to try to piece together the green columns - which is very difficult for human cognition.
Let’s adjust these defaults and make over this chart so the story pops out.
Change the title to state the key takeaway
Truly, the simplest way to tell a story with your data is to maximize your chart’s title space to, well, tell the story.
Rather than the vague “Click Through Rates by Email Campaign Type” use your insider knowledge and your data literacy skills to interpret the data and highlight what you see: “Click through rates for advocacy emails gained ground while fundraising lagged behind.”
The title is now a complete sentence that tells the audience exactly what you want them to notice about the email campaign data.
Now that we have a clear story to highlight, we need to reformat the rest of the chart so the evidence that supports that story is more obvious.
Also read: Seven ways to build an equitable website
Use color intentionally to highlight the story
Bold colors on every data series will create visual competition. Once you see the story, reserve your bold colors for those data points and tone down the other data with gray.
Gray will fade the rest of the data into the background and let the more important data points jump out.
When you apply action colors to the salient points in your chart, your audience knows exactly where to look.
Even better, leverage that storytelling title and color-match the words that correspond to their associated data series. Now you have a chart title and a dataset that unite. But we can make this chart even clearer.
Simplify axes and labels
Despite the smart changes we’ve made so far, viewers can still get bogged down by the redundancy and clutter in the graph. And it isn’t your fault! Every software program we use to make charts will include clutter by default. You just need to learn to recognize the clutter and cut it out.
Look for repetition that doesn’t add clarity. For example, we don’t need CTR in every label along the x-axis if the title is already telling everyone that this data is about click-through rates. We can ditch the (%) all along the x-axis AND the y-axis label “percent” if we just add percentage signs to the data labels.
In fact, if we have data labels on every column, we don’t need the y-axis at all. Bye!
Clutter also shows up as excessive and unhelpful lines.
If we remove the y-axis, we don’t need its associated horizontal gridlines either. Also, you can almost always remove your chart border.
These might seem like minor tweaks - and they are so minor it will only take you a moment to implement them - but they add up to a chart that’s much easier to read.
We could stop right here and we’d have a suitable chart that does the data justice and sparks conversations with our team. But let’s go one step further.
Try a different chart type
Though we have significantly improved this visual, the chart type itself is still posing a problem. To track progress on any one email campaign, our brains have to weed through and attempt to ignore the data for the other email campaigns. This dilemma is inherent to every clustered column chart.
Convert to a line chart.
Line charts do a better job of showing trends over time. They connect the data points for each series so our brains don’t have to work overtime to see where we increased sharply or where we stayed flat.
Because the inclines are now so clear, I took a risk and removed the exact data label from most of the data points and only retained it on the most recent time period. I assumed the most recent quarter would be of most interest to my audience, but you know your audience and what they’ll want to see.
The remaining data label was also an opportunity for me to embed the legend directly into the chart. No more attention bouncing back and forth between the legend entry and its corresponding line. Saving everyone’s mental energy is always a win.
The story told in the chart’s title is now much more obvious in the chart itself.
Also read: What 'user-centered' could mean for you
Bottom line
You are the person best positioned to know the story in your data. You just need to adjust the default chart formatting so the story you see is just as evident to everyone else.
- Use the chart title space to articulate your point as a complete sentence.
- Gray out the less germane data and apply an action color on the data you want your audience to notice.
- Strip out clutter like redundant labeling and unnecessary lines.
- Explore alternative chart types that can better highlight your data.
With just a few moments of effort (that will get even faster with practice), you will end up with a chart that produces clearer insights, easier decision-making, better engagement, and tells your data story effectively.
Stephanie Evergreen
she/her/hers
data visualization consultant at Evergreen Data,
Stephanie Evergreen Ph.D. is the author of three data visualization books and a free class on how to build a culture of data visualization at your organization 👉🏼 BuildADataCulture.com
