Join Doug Rose for an in-depth discussion in this video Reporting isn't telling, part of Learning Data Science: Tell Stories With Data.
- Business presentations can be boring. They're not really set up to be interesting. Most of them communicate the status of your work. They're like a verbal 'reply all' to your audience. That's usually fine for a typical status meeting. You'll want your data science teams doing something different. Remember that data science is about using the scientific method. Your teams will be going through lots of questions. You'll be panning for gold looking through your data for key insights. You'll need to explain a lot of different outcomes.
Almost all of your challenges will be about explaining what you find. Your data science team will've explored the data on their own. Now, it's time to create the stories that explain what it all means. I once worked with a data science team that was focused on delivering promotions to credit card customers. The team asked a bunch of great questions. Some of them were about what their customers were purchasing. One of the questions led to a key insight. They wondered are customers accepting promotions in bunches.
They found that if a customer accepted one promotion, they would be more likely to accept the next bunch of promotions. This increased what they purchased and that helped the credit card company. They wanted to show this insight at one of their storytelling sessions. The research lead came up with a presentation but it wasn't a story. She just wanted to explain the findings. It didn't make any attempt to explain what it all meant. She just wanted to point out that customers were more likely to accept promotions in bunches. Then she'd leave it up to the room to decide what to do with that information.
I reminded the research lead that the storytelling session is not the right place to just present the data. They needed to weave together an interesting story so the room will be engaged and connect with the insight. So I asked the research lead what she thought was happening. She said that the data suggested that most consumers got money in batches. It makes them spend during up times and save during down times. Customers were also more likely to accept promotions depending on different phases of their lives.
We both agreed that this was a much better way to convey this insight. Everyone in the room would be able to share in this experience. They all had less or more money during different phases of their lives. Why not use this shared experience to convey the insight about how the customers are accepting promotions? The research lead came up with a new presentation. They created a story called the impact of promotions during different phases of people's lives. She opened up her story by relating an anecdote from her past.
She said that when she was in college she had a roommate. They used to receive coupons in the mail that offered two-for-one meals. After class, her roommate would come home and ask what was in the mail. That way, she could see where they were going for dinner that night. After four years of college, they ended up having the same tastes in food. A few people laughed and they were already starting to think about how people use business promotions during different phases of their lives. It was a much easier transition for the research lead to present the story.
She described how most customers will accept promotions at different rates during different phases of their lives. That instead of a steady stream, these customers will accept promotions in bunches. The anecdote combined with the story did a lot to encourage the audience to participate. A lot of them became very curious. They asked questions like, does that mean that a promotion might be more effective if someone feels like they're in a vulnerable phase of their life? Another person asked, how far the team was from predicting when a customer might be entering a phase when they would accept a lot of promotions? If she'd just given a typical presentation, she would never have this level of engagement.
The story drew them in and helped them connect the insight to their own experience. The audience thought of times when they might have been more likely to accept a promotion. Then they were able to ask interesting questions and build on this insight.
- Structuring a data science story
- Defining plot, conflict, and details
- Going beyond reporting
- Knowing your audience
- Working with data
- Introducing visuals
- Eliminating distractions
- Incorporating metaphors
- Motivating the audience
- Avoiding pitfalls