Getting your audience’s attention, keeping it, and persuading listeners of your point are all hard to do in a world where most listeners start out thinking, and feeling, “I’ve got my own scheisse to do.” John Weathington’s recent post in Tech Republic, “Be the Hemingway of Data Science Storytelling,” makes the point that presenting data, which can be dry, is more effective if it incorporates elements of story – a protagonist, a journey with challenges, and a conclusion. Jeff Leek’s “The Elements of Data Analytic Style” has a chapter about presenting data that emphasizes story as the method for communicating results.
Great points. Essential.
But how literally should “story” be taken? Story, often romanticized as an abstract concept by omitting an article in front of it, can be an enchanting idea.
But as a concept it is sometimes more intuitive than descriptive, and it is hard to apply effectively. Thinking it means we should always communicate data science as if narrating a linear plot from the first-person perspective would be a mistake. So how can we use story effectively to capture attention, keep it, and persuade our listeners?
This post, focusing on audience and empathy, is the first of three posts exploring how to implement story when presenting data.
First, think carefully about the audience. Then Empathize.
Identify and stratify your audience, to the extent possible, in order to understand what matters to them. Then, prepare the story your audience will relate to, which may or may not be the story you experienced. Connect to your audience’s APIs, instead of documenting your own.
My colleague, Dave King, recently posted about a data science project on global public health that Exaptive is doing for an NGO. We are passionate about how our platform is facilitating the application of algorithms from one field within another, based on collaboration between experts from different fields, and we want to find other likeminded data practitioners. So for our blog, the story’s main protagonist is the guy with the algorithms, the data scientist, and his journey working on the data in our platform. It’s a great story.
If, however, Dave were presenting results to an audience of researchers, the people who generate data, that would probably not be the lead. The protagonist might be the head researcher aggregating data from dozens of other researchers and her winding journey from disparate data to actionable insight. That’s also a great story, about the same project, but it would sound, and look, quite different.
If you’re using story to describe a project and its results, your choice of protagonist, journey, challenges, and conclusion can change dramatically depending on the audience. This may seem obvious, but it’s deceptively hard to do with your story as you know to do with data, change perspectives. (Most authors write semi-autobiographically first, right?)
After immersing yourself in your work, it takes self-awareness, empathy, and intellectual agility to shift perspective and tell your story to communicate effectively with different audiences. Take on the challenge, though, and you’ll rise above the din.
When in doubt, to thine own self be true.
After all this empathizing, reconnect with your expertise and passion and make sure you still shines through. Unless you’re an actor (and a good one), good communication cannot be faked. Being yourself and communicating what you know well, while meeting the audience where they are, is what’s effective. The perspective you employ may change, but the reality you are describing should not.
Nailing it . . .
This post on Nate Silver’s fivethirtyeight.com about predicting someone’s age by their name is an impressive example of balancing story, audience, and truth. The piece is solidly rooted in the data and analysis, explaining the facts and providing satisfying visuals. The story, though, is “you” going from name to name, exploring if and how patterns in name popularity can be used to guess someone’s age.
Concrete techniques in the next post . . .
Audience and empathy is where I start, but there is plenty more to think about when it comes to communicating data science effectively. Stay tuned for:
Comments