Thank you to everyone who attended my session at the Enertia User Conference in Las Vegas earlier this month! It was a blast meeting everyone, and I can’t wait to see your BI solutions come to life. Big shout out to my dad, Steve Hughes, for introducing me to this amazingly talented group of people, and to Fernando Salazar, for welcoming me into the Enertia community. Below is all the research used in my presentation on Storytelling with Data. Please use this link for the GitHub folder containing the presentation, sample data used, and a sample report that we worked on together. Happy coding!
In 2015, the average company spent $7.4 million (USD) a year on data-driven initiatives (IDG Big Data & Analytics Survey). In 2021, the average company expected to spend $12.3 million (USD) on data initiatives in 2022 and 55% expect their budgets to increase (Data & Analytics Study). Data and analytics budgets have grown nearly 40% in five years, yet Foundry found in their 2022 study that “lack of appropriate skill sets is a top challenge” (2022 Study). It’s clear that data is a priority to many organizations, but without the proper training on how to use data to deliver insights that data sit like a box of nails without a hammer.
To turn our millions of dollars worth of data into actionable insights, we need to tell a story with it. Why a story? Storytelling has been and is a vital way humans understand and share the world around them. If numbers were as interesting and memorable as stories, our classic works of art would be the Pythagorean theorem instead of the Iliad and Odyssey. Listening to a story engages multiple parts of the brain, which is why listening to your great aunt’s story for hours is exhausting, but also why stories are more memorable. Here are the parts of the brain engaged when you hear a story:
- Wernicke’s area (language)
- Amygdala (emotions)
- Minor neurons (empathizing)
The hippocampus (short-term memory storage) is more likely to push stories into long-term memory than numbers because they engage multiple areas of your brain. This is why so many math classes include word problems. Thankfully, visualizing your data doesn’t need to be like a math test.
There are a few main elements to be successful at data storytelling:
Data powers the visualizations that can communicate your story clearly and memorably. Similar to a picture book, a good report will quickly communicate a story and a great report will help you change that story through actionable insights. So how do we go from numbers to a story? From math class to lit? Through connecting visuals and numbers to elements of a story. Every story contains a few common elements: characters, setting, conflict, and resolution.
For example, imagine you have seen a downward trend in sales from Texas for smart cars. You characters would be your previous and current customers in Texas, setting is the timeline of the downward trend and the customers across the country, conflict may be that news stations in Texas ran a story about hackers getting into smart cars, and resolution may be to advertise a security patch through local news stations in Texas. There may not be a conflict in your story (maybe you’re exceeding your goals), so that section can be skipped in favor of focusing on what is going well.
Always remember that applying analytical techniques to managerial problems requires both art and science.Jan Hammond, Harvard Business School Professor
Creating Data-Driven Decisions
Storytelling with data can leave an audience asking questions such as “how do we make our sales go up?” or “what is xyz branch doing to increase profit percentages higher than everyone else?” This is where the resolution portion of the story comes into play. A lot of good reports will “tell the news”, that is they will display a current state of affairs retroactively. To generate a true return on investment (ROI) from our data, we need to use it to proactively drive decisions.
For example, in 2008 Starbucks closed hundreds of locations. Howard Schultz returned as CEO and declared they would use data to place their stores strategically going forward. Starbucks now consults with an analytics company called Esri to analyze retail locations for various variables that are proven to drive coffee shop traffic and overall success (reference).
The key to building a data-driven culture is to avoid deceiving your audience. Visualizations are a powerful tool, but they can be used to trick the viewer into disproportionate understandings of data. For example, New York Times came out with a visual in 1983 that showed the mandated fuel economy standards changing from 18 to 27.5 miles, an increase of 53%. The graph in the article, however, had an increase of 783%! A good way to build trust in your data is to avoid the Lie Factor.
The representation of numbers, as physically measured on the surface of the graphic itself, should be directly proportional to the quantities represented. The Lie Factor = the ratio of the size of an effect show in the graphic to the size of the effect in the data.Tufte
This all sounds well and good, but how much money can using data instead of your gut actually save you? According to a study by Harvard Business Review, Fortune 1000 executives have seen the most value by aiming to decrease expenses. The second highest is by finding new innovation avenues. Decreasing expenses is great, it allows your company to lower overhead costs of operation and increase profit margin. However, there is only so much overhead you can eliminate which is why finding new innovation avenues will be the long term ROI from using big data.
Building a Data-Driven Culture
One of the biggest blockers to using big data effectively is adoption by the larger business. According to HBR, more than half of Americans depend on their gut to make decisions even if there is evidence that disproves their theory meanwhile data-driven organizations are 3x more likely to report significant improvements in decision-making. There are many ways to become more data-driven, even in your daily life. HBR summaries a few easy steps to take to begin building a culture of cultivating insights from big data:
- Look for patterns everywhere
- Tie every decision back to data
- Visualize the meaning behind data
- Consider furthering your education to learn more data analyst strategies
Thankfully, this approach does not need to be all or nothing. It can be as simple as looking for patterns in your personal life like spending more money on ice cream deliveries after watching sad movies. A simple cost-saving would be to stock up on ice cream in advance and avoid the delivery fees. Finding a new innovation source may look like watching romantic comedies instead of sad movies or switching to popcorn which is much more cost effective.