The Importance of Storytelling

What’s the big deal about ‘Stories’ in Analytics?

Ruth Guthrie, Cal Poly Pomona

If you’ve read trade magazines and websites about data analytics, you’ve probably seen a lot of attention given to “data driven story telling”, “letting the data tell the story” or “Story telling with data”.  Even Tableau has a tab at the lower right for creating a “New Story”.

•  The Best Data Scientist Know How to Tell Stories (HBR, 2015)
•   Use Data and Analytics to Tell a Story (Gartner, 2018)
•  How Do You Tell a Story With Data Visualization? (Forbes, 2019)

Why Storytelling?

Storytelling is inherently friendly.  Most university students have a tremendous fear of statistical analysis.  Statistics is a course with one of the highest fail rates at most universities.  It takes a little work to understand your ANOVA table, p-values, and Pearson’s R.  Statistics and data analysis is often mystifying to people. Attaching visualization and storytelling to data analytics makes it more attractive.  Being able to explain what the data is saying, without losing your audience in the numbers, makes the data understandable in a highly concise, comprehendible, memorable way.  Better yet, provide a dashboard so they can easily play with the data to confirm or disprove their ideas.

Storytelling is a great buzzword.  Data analytics is one of the hottest growing programs at Universities today.  Storytelling makes a great title and perhaps a great learning outcome. It sounds like we’re actually doing something, not just pulling red and green balls out of a sack.

Storytelling puts the focus on making information clear and actionable.  All the data in the world isn’t helpful unless you give it context and direction.  If you cannot tell the story of what the data provides and why that is interesting, you might miss an opportunity to provide an analysis that results in decision making.

What’s the downside?

Keith McNulty (July 21, 2018) Beware of ‘storytelling’ in data and analytics, cautions that storytelling may have the impact of focusing more on the story than the research-based analysis necessary to truly examine a problem.  The risk is that a client/boss would be persuaded to act on something that was more visual driven than data driven. Sometimes the data is bad or there isn’t enough of it to draw a conclusion.  Further, if data professionals are pressured to produce a result that support specific initiatives, then the integrity of the analysis is compromised.

We want our stories to be novel and interesting.  Likewise, there is a pressure and hope for data analysis to come up with a ‘significant result’.  As we build data science curriculum, the analysis, ethics, asking the right questions and storytelling will all play into effective problem solving, even if the results are not exactly as we expected.