Before falling in love with statistics and analytics, I traveled a long path to avoid numbers in my life. Sounds ironic but it's true. I was petrified of them. Through my early school years and college life, I keep them at arm’s length as much as possible. Yes, I took Finance and Accounting to attain my degree in Marketing, but that was more akin to torture than enjoyment.
It wasn’t until I entered Graduate School to pursue my advanced degree in Communications, that I faced statistics front and center. Since I was so determined to make it work, not only I embraced them, but also took all the advanced statistical courses I could get my hands on during those semesters.
That’s when I learned to marry communications with data.
I’ve said it before, that data is an aggregation of stories, so for me, it was very easy to bridge the two.
However, this is not the case for most people I know. So I would like to share three learnings from my experience communicating quantifiable information. Some of these were given to me by mentors or gained along the way by practicing and reading about the topic. They are simple to follow and easy to replicate. Here we go!
FOCUS ON YOUR MESSAGE
You will have to determine what among the whole jumble of data that you inherited from your boss or that resulted from your research efforts makes sense to advance your proposal. Focus on a particular point and define it as an actionable headline. That is your insight, the piece of information that gives you and your team the aha! Moment. That’s what you have to feature on your charts, not generic titles!
For example, compare these two:
Gross Earnings for 2016
Energy Costs Consume Bulk of Gross Earnings during the past year.
Can you see the difference? What was your reaction to the first title? If it sounds like something you would have written, I know exactly the feeling. I composed those generic titles too, and many.
Now, how was the feeling with the second title? More engaged? It gives the audience a sense of what is being discussed and provides for a richer exchange of ideas. Some might even question your assessment, but that is better than just glazing through your slides without any contribution to advance the discussion.
Another relevant point regarding this strategy is to use strategic focus for your charts. Once you identified that story and point of interest in your charts, use design techniques to bring the attention of the audience to that point. Otherwise, they might get lost.
You can also do this by utilizing the Gestalt Laws. The Gestalt Laws are psychology grouping principles that related to the concepts of Continuity, Closure, Connectedness, Proximity and Similarity. In essence, by using these principles, you will enhance the meaning of your visuals by augmenting or reducing the perception of your audience as it relates to the content inside your charts.
By decluttering your charts, you will also be able to attain increased focus and clarity. You declutter your charts by removing any repetitive labels, gridlines or unnecessary information that does not directly support your core message.
To quote presentation expert Garr Reynolds:
“Simplify, to amplify.”
CHOOSE AN EFFECTIVE CHART
This proposal might sound like its common sense, but you will be surprised as to how many people make a mistake choosing the wrong visualization to express their quantitative information.
Here I’m going to address the three most common mistakes in choosing a chart.
1) Our Friend the Pie Chart
Yes, it’s common.
Yes, it’s easy to read (sometimes)
We’ve known her for a long time and feel at ease with her, but this chart has a dark side.
Have you tried to compare two pie charts next to one another? Isn’t it like a ping-pong exercise where you have to go from one chart to the other in order to make some sense of the data?
It doesn’t make sense to put your audience through that. Do yourself and them a favor, unless you are comparing same category items as percentages from a whole; move on.
Nowadays this is the case for most comparisons, so be ready to amicably part ways.
2) The Underused Bar Chart
Bar charts are some of the most practical and versatile visuals that you can muster since they are very easy to create, read and people are very familiarized with them. Unfortunately, this same familiarity deters some to use them and pass over to choose other fancy but less applicable visuals instead because of their inclination to think of bar charts as boring. The trick about bar charts is their arrangement depending on the type of comparison you must convey.
If you are to compare items as a ranking, such as the sales personnel performance during the second quarter; then the vertical alignment is the most appropriate for two reasons. The first one is that it more clearly communicates rank by avoiding confusion in your audience whom might think of it as a time series comparison as it would be the case in a horizontal alignment.
The second reason is more related to good design practices. A vertical alignment will more likely accommodate longer labels rather than a horizontal alignment, where these labels would add too much noise to the chart.
3) The Spaghetti Chart
We all have had the misfortune of seeing these at least once. Spaghetti charts can be easily avoided, but they keep popping up their ugly heads in presentations daily. I hypothesize that the reason for this is due to the pressures of keeping presentation decks slim and concise. This pressure creates an undue burden on presenters, that to reduce the number of slides, chose to jumble up all competitors or products from a category together rather than separating them into more meaningful comparisons.
Spaghetti charts are meant to evaluate time series trends, but they barely accomplish anything beyond confusing your audience.
Instead of falling into this trap, choose to separate into simpler representations that communicate your message more effectively. I prefer to separate them using a quadrant method, in which the company’ s brand compares against the top competitor in each region (East, West, North and South) or other similar categorizations.
KEEP IT HONEST
Integrity is an essential part of communicating with words or numbers. Shall you choose to be creative with your chart, all it takes is a discerning audience member to point out inconsistent or deceiving representations of the data to throw your credibility out the window.
To avoid this mishap, stick to two practices.
One, Keep a zero baseline.
Two, beware of making facts or assertions out of small sample sizes.
A zero baseline is important because it allows for a clear and unbiased representation of the data. I get it, that as communicators we must focus on a message to convey and that sometimes that message can be accentuated and placed into context creatively and truthfully, so that it delivers a powerful impact on the audience. However, a risk with changing the axis is the potential confusion it can cause to the public if not placed into the appropriate context. That confusion can lead to a loss of credibility. In my opinion, is not worth it.
The second practice is commonly ignored by individuals who not familiarized with data collection or analysis. However, ignorance is not a justification for this mistake.
A sample is meant to be a representation of the larger population. Since we can rarely collect information from a whole population, we use samples. Samples need to be representative and have a significant size to reduce their margin of error. Making assumptions on small sample sets poses a considerable burden on the accuracy of your conclusions and future recommendations. When providing findings to your audience based on sample sizes, always pair them with their corresponding margin of error so that you can more accurately convey your message.
From how to focus your message, to choosing your optimal visualization medium to build charts correctly to avoid confusion among your audience – simplicity is king.
If you find that the production of a chart is becoming too hard to read or to create, take a step back and rethink your approach.
Instead think: How can I make this simpler?