From predicting the weather to anticipating disease outbreaks, computers enable us to gain useful insights from massive data sets. However, while a skilled analyst can scan a sheet of numbers and spot the important trends, most people do better when data is presented within a simple, logical, compelling narrative.
Studies have found it’s 20-22% easier for people to remember information they learned from stories, and an analysis of popular TED talks found the highest-rated speakers spend 65% of their time telling narratives, versus transmitting abstract information.
If your job requires you to present complex data to lay audiences, here are some quick tips for telling an effective “data story”:
- Answer one question. To make sure your narrative stays focused and easy to follow, ask a straightforward question, then tell a story supported by data to answer it. For example: “If we keep eating so much sushi, how soon will bluefin tuna go extinct?” or “If we raised the prices by a dollar, how will customers react?” or “What is the cost to society of not letting young girls attend school?”
- Tie data back to reality. Don’t assume that your audience will be able to appreciate the significance of a statistic on its own. Make sure to regularly highlight the real-world implications. For example: “Four additional centimeters of rainfall is enough to flood sewer systems and water treatment plants, turning our drinking water cloudy or brown.”.
- Use text, tables and graphs – but not all at once. Tables with numbers are good if you want to invite the audience to examine the source data and walk them through your team’s analysis. Charts help audiences quickly arrive at an insight without having to comb through the numbers (just be careful that your visualizations aren’t misleading). Text bullets are a quick way to distill key insight without getting into the analysis.
Just be sure to keep your visuals clean and uncluttered, ideally with one graph with large, easily legible labels conveying a single idea per slide.
- Tell the whole truth. Anyone who works with data knows that, behind the nice, clean graphs and tidy tables, the work of collecting and interpreting data can get extremely messy. And while we don’t want to bog down an engaging story with long disclaimers, we have an obligation to let the audience know – in terms they will understand – how we got our data, how we analyzed it, and if there’s anything in our data that might complicate the narrative we’re weaving.
As a rule of thumb, if an assertion you’re making could be contested, say so at the time you make it. If it merely requires clarification – it’s OK to save the caveats for the end.