Over the last few years, marketers have increasingly turned to data visualization to report on trends and campaign performance. Charts, graphs, and other visual aids help break down complex analytics into digestible chunks, giving the team quick visual insight into the successes or failures of marketing efforts so you can take decisive action.
Too often though, data visualization has the opposite effect — instead of empowering your team to take the right action, a misguided chart or meaningless graph can confuse, bore, and alienate your audience.
Here’s our guide on how you can go beyond simple data visualization so that you can represent your information in a way that’s understandable, engaging, useful, and (most importantly), actionable.
1. Identify Actionable Information First
We’ve all sat through a a meeting where some bad news was reported — a chart shows the line of web traffic or sales sinking instead of heading “up and to the right” — and have walked away without any information that seems relevant or arms us with actual insight into what to do next to solve the problem at hand.
It’s crucial — and will save you precious time — to only do analysis that actually moves the needle on your marketing. At Revere, we believe that statistical anomalies and outliers — an odd bump in sales, a drop in spending — are especially actionable where they can also serve as a window into everyday marketing decisions (SEO, advertising spend, webpage tweaks, etc.) that truly drive growth or failure.
“Only visualize the data that actually moves the needle on your marketing success.”
Here are just a few examples of meaningful data events. You don’t need a degree in data science to understand these are significant to your business:
- Spikes and dips in advertising spend within a given time period
- Significant drops and increases in click-through-rates
- Unusual spikes and dips in sessions and conversions
- Money lost on underperforming search terms
- Cost-per-conversion on underperforming ads
- Significant increases or decreases in overall revenue
- Traffic sources that drove unusually high revenue on November 14th
These aren’t just stats — they’re discrete and important events that will help you compare user behaviors, contextualize them in terms of your specific marketing efforts, and take decisive, specific action. (Hint hint – Revere automatically and regularly sends you this kind of significant data.) Once you’re armed with this information (and not before), the next step — representing your insights with a clear, visual data story — will make an actual impact.
2. Choose a Story to Tell
You’ve got your insights chosen, and you’re ready to represent them—fantastic! Now it’s time to a) choose the story you want to tell to capture your audience and b) pick the visual format that best expresses that story.
We can call pie charts, word clouds, and other visual analytics diagrams “data stories” because they act as (ideally, actionable) narratives or messages you tell about your company. The story could be something like “When we added more Free Trial buttons to our site in January, conversions spiked and our overall revenue went up astronomically!” (We all like that story).
In order to get through to your audience, you need to choose the appropriate vehicle (usually a chart) for that story. With few exceptions, you’ll be representing your data in the following ways:
A) As a composition: This is used to gather different pieces of information that make up one collective. You might, for instance, tell a story about the distinct navigation pathways your site visitors used to get to your site (search engines, links, etc.).
Chart to use: Pie chart
image source: kaout.org
B) As a comparison: As it sounds, this is used to compare sets of variables, like how many people visit landing page X vs. landing page Y within a given time period, or to compare advertising revenue by quarter.
Chart to use: Column Chart
C) As a relationship: An attempt to show a connection between several variables. For instance, if we choose to visualize the data events mentioned earlier, you might attempt to show the relationship between advertising spend and significant bumps in traffic or conversions. As you might guess, we believe that this is one of the more actionable approaches to data and visualization.
Charts to use: Bubble Chart, Scatter Chart
image source: wordstream.com
D) As a distribution: This is used when it’s not entirely clear if or what the relationship is between two or more pieces of information, like between Facebook advertising spend (or, in the example below, organic search ranking) and an increase in click-through-rates.
Chart to use: Scatter Chart
Before settling on one visualization, try comparing different ways to present your data—apps like Tableau can help you do this quickly and succinctly, and Google Spreadsheets or Excel are reliably great standbys.
Check out this infographic and blog from infogram.com for deeper info on different chart types and how to choose the right kind for your data set.
3. Make Your Story Interesting
Even if you’ve picked the right data to represent, and even if you’ve chosen the right vehicle for your data story, none of that hard work matters if your audience is bored out of its collective mind.
Humans are hardwired to learn via engaging, interesting, and sometimes humorous narrative — it’s why our life lessons come stored in bizarro tales about frogs and hares (slow and steady wins the race!) and buck-naked emperors (power doesn’t always equal competency!).
A story — even an educational data story — must, at the least, have a protagonist (perhaps your marketing team, who is obligated to sit through your presentation), a clearly represented challenge or obstacle, and a little bit of humor (that emperor again).
Another way to put it: your team needs to feel both a personal stake in your data and to be entertained to some degree to stay engaged and learn (whether this is an insult or compliment to the human mind, we’re not sure). Information should be “immediate, personal, and local”.
“Your team should feel a personal stake in the data you present.”
To make your data story more personal, include a narrative about how you or a colleague relate to the data, or narrate the data in a way that truly situates your department as the embattled protagonist trying to slay the dragon of high bounce rates (for example). And on that note, a little humor, too, can go a long way. Cheesy? Yes. Engaging? Definitely.
A good story, like a good old folk tale, has a lesson toward the end. This is another way of reiterating our original point: make your data (your story) actionable. What will those hearing the story learn toward the end? What can they take to make decisions in the future? Including this kind of narrative guidance is the difference between something that’s useful and engaging and something that’s useless and yawn-inducing.
4. Make Your Story Beautiful
As the cliche goes, humans are drawn to pretty things like moths to a flame. But it’s not just about pretty: contrasting colors, legible language, and other principles of design help us focus.
How many times have you looked at a table or graph and had to squint just to read the 15 items being represented? Giving your audience an extra challenge to accessing information can cost you your audience’s attention and credibility.
Some tips to designing a great chart, graph, or diagram:
- Skip the fancy formatting and effects — they’re distracting.
- Make effective use of chart titles
- Keep it simple—if there’s too much data in a single chart, consider splitting your one chart into several smaller charts.
- Don’t be afraid not to visualize data — although we’re giving you tips for great data visualization in this post, sometimes things are better left said (that is, written).
5. Make Your Story True
It shouldn’t have to be said, but honesty isn’t just a great moral principle—it’s also crucial to your credibility and to the usefulness of your information.
As the Harvard Business Review puts it, “A visualization should be devoid of bias. Even if it is arguing to influence, it should be based upon what the data says–not what you want it to say.” Bias and also unintentional inconsistencies, the author suggests, can compromise your credibility.
Intentionally leaving out pieces of data that don’t adhere to you narrative is also known as a “lie by omission.” If the data doesn’t fit your story— revise your story, not your data. Your team isn’t looking for a data spin doctor, but someone who can help illustrate what matters.
“If the data doesn’t fit your story— revise your story, not your data.”
6. Stay humble and share the wealth
Finally, don’t keep that good data to yourself. Share your visual data story with anyone that it might help (think of it as like a news or status update). After all, cross-functional collaboration is an important part of a healthy and functioning business.
Being transparent and sharing your work is a benevolent act—you never know when the sales team, for instance, could use your insight. It’s also a (positively) selfish act—other team members and departments might read your data differently, tell a more insightful story that you had missed, or simply catch errors. This exchange of information can have a salutary impact on your company’s potential for success.
Revere was created to allow for this kind of important data sharing. Teams can integrate with Slack and get notified in real time when their marketing performance changes (unusually low CTR, dramatically high cost-per-conversion, etc.) in a way that’s meaningful.
Whether it’s via a beautiful chart, a compelling graph, or a conversation about a Revere alert over Slack (“Woah! That bounce rate yesterday, though? What’s that about?! Let’s chart it!”), great marketers know how to share the data stories that matter most.