Join Doug Rose for an in-depth discussion in this video Tell an interesting story, part of Learning Data Science: Using Agile Methodology.
- There's a big difference between presenting data and telling a story. For one thing, telling a story is much more challenging. It requires a lot more work. You're combining the data with what you know about the business. Then you're relating your insights to what you know about the world. When you put up a data PowerPoint presentation, you're saying, "Here's what I see." When you're telling a story, you're saying, "Here's what I believe." That's a much more difficult thing to do. In a way it's also more personal. That's why storytelling is such a valuable skill.
In a modern office, almost anyone can create a chart using Microsoft Excel. People think of this in the same way as knowing how to use a computer. When you're telling a story, you're doing several things at once. The first thing you're doing is synthesizing much of the complexity in your data. You're explaining something that's complex in a simpler way. You're also defining the motivations of the people who are involved in creating this data. The second thing you're doing is bringing your own knowledge about the organization.
This could be through your experience or even some research. You're taking a simple observation about the people and the data, and then putting it in the context of the organization. You're not just using data to talk about what and where. Instead you're presenting about the why. The third thing you're doing is making your data more memorable. Several studies show that when you present a PowerPoint slide, very little of the information gets through to your audience. Those bullet points might be easy to create, but they're also very easy to forget.
A story has an easier time capturing your audience's attention. You can weave together a good story and try to get everyone engaged. Finally, a good story has a call to action. It will either tell you something new, or it will justify your continuing to look. You can even include your audience in the exploration. Let's go back to our running shoe website. Imagine that your data science team has been working on a question about increasing sales. You work with the team to break down the question into smaller questions.
One of these smaller questions was, are people buying things on their wish list? The research lead and the data analyst worked together to create a quick and dirty report. They want to see how many of their customers' wish list items were converted to purchases. Then they create a time series to see if these purchases were going up or down. Typically the team will get together in a meeting about visualization the day before the storytelling session. It's in this meeting that they'll try to convert the raw data in simple reports into a nice visualization.
They'll use this to tell an interesting story. The data shows that in the summer months, people are more likely to convert their wish list into purchases. That's just the raw data, but it's not a very interesting story. The real question is, why are people waiting until the summer to purchase items they wanted in the winter? The data science team decides to create a visualization. They use the title, Summer Dreamers: Why Winter Shoppers Buy Shoes in the Summer. Then the data analysts use a whiteboard to come up with a first draft of the data visualization.
Notice how a story already starts to make the data more interesting. Imagine if the data analyst used the title, Annual Wishlist Conversion Rates. Then they included a simple time series on a PowerPoint slide. Something like that wouldn't grab anybody's interest. There's no context, or a call to action. The next day, the data science team used their Summer Dreamers visualizations to tell an interesting story. Many of your customers are thinking about running in the winter, but they're only buying shoes in the summer. Notice how the story encourages further questions.
Are people in the winter running in old shoes? Maybe they're just not running. Do they not need new shoes because they're mostly running indoors? Should we make a special running shoe designed for indoor running? Hopefully you'll get these questions during the storytelling session. Then you can add them to your question board. If you can tell a good story, then everyone in your organization will want to take part in your team's discovery.
This course shows how to structure your work within a two-week sprint. See how to work within a data science life cycle (DSLC)—a methodology for cycling through questions, research, and reporting every two weeks. Explore key practices to help your team break down the work so it fits within a two-week sprint. Learn how to use tools like question boards to encourage discussion and find essential questions. And most importantly, learn how to grow your team's shared knowledge and avoid common pitfalls.
- Defining data science success
- Determining project challenges and criteria for success
- Using a DSLC
- Iterating through DSLC sprints
- Creating a question board
- Breaking down your work
- Adding to organizational knowledge
- Avoiding pitfalls