Join Doug Rose for an in-depth discussion in this video Eliminate distractions, part of Learning Data Science: Tell Stories With Data.
- When you're telling your data science story, you'll be tempted to share a lot of data. Some teams think this is a good opportunity to show off their work. Unfortunately, the story telling session is not a good time to show the ins-and-outs of data science. As you've seen, a good story telling session will have the data in the background. It will be a supporting actor in your story and not the main character. That's why one of the key things you want to remember is that you need to strip away anything that doesn't enhance the story.
As you get closer to your session, you want to spend more time eliminating than adding from your visuals. There are two places where you want to eliminate as much as possible. First, you want to make sure that your data visualizations are as clean as possible. Second, you want to make sure that there's just enough characters, plot, and conflict to hold the story together. When you get close to your session, you should be taking things away. You want to strip away everything that doesn't get to the very essence of your story.
At the same time, you want just enough details to keep things memorable. I once worked for an organization that had story telling sessions that were driven by one of the directors. He started the session by congratulating the team. Then he would go through how important it is to the company to become more data driven. He would go on to say that it was a big part of their larger organizational strategy. After ten minutes, the director would start telling the story. Unfortunately, by this time, a lot of the audience stopped paying attention.
Some of the executives would start to look at their smartphones. Others would just stare blankly at the first slot. The extra information at the beginning was too distracting for the audience. Now imagine if the director immediately engaged the audience. What if we said we think we found a way to better predict our customer's behavior? By looking at patterns, we're getting better at telling what our customer will buy before they even think of buying. Then he can start the story about a typical customer.
Either way, he's engaged the audience from the moment he starts telling the story. He's eliminating the distractions and zooming right into the value. Remember that the audience wants to take something away from your story. As soon as you start your story, you should give them something interesting. As you continue to give them more, they will be drawn deeper and deeper into your story. That's why you should work to strip away all the organizational norms that are usually the start of every meeting. A story telling session is a special occasion.
You should never treat it like a typical status meeting. You don't need to congratulate the team. You don't have to show the importance of their work. When you eliminate these distractions, the audience is more likely to attach meaning to your story. You also shouldn't need many data visualizations for your story. When you use them, try to eliminate the distracting information. Pull out any detailed data that could easily be summarized. Remove any text that you don't need from the image. When you look at the data visualization, you should ask yourself if there's anything that could be stripped away.
If you remove it, will it take away from the story? Not every point needs a text label. Not every series needs to be written out. If you have a time series that shows Monday through Friday, not every day needs an abbreviation. When you strip away these distractions, then you're doing some of the audience's work for them. They don't have to think as much about the visualization. They don't have to wonder about what you're going to talk about. You can eliminate these obstacles between you and the audience. If you find yourself adding more information as you get closer to your story, then you're probably making it too complex.
The most important thing to remember is that less is more. A good story telling session isn't like a firework show. You don't want to dazzle the audience with sights, sounds, and swirls of color. It should be simple and focused. There should be easy to read visualizations in a simple story line with just a few memorable details. The fewer distractions you have, the more your audience will take away from your story.
- 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