Framework for Storytelling with Data

Data by itself doesn’t help us very much. We can gather as much of it as we want, but it won’t bring people together or get them to do anything. Because data has so much to offer, when we look at our data, figure out what it means, and explain what we’ve learned in a way that everyone can understand, we’ve turned it into something very valuable: a story.

Stories are more interesting than facts because they stay with us longer. The same goes for giving information to our team and to the people in charge. Data storytelling can help people act on the things they learn from data. Without good communication, our audience might not notice or remember our insights. Hard and soft skills together help us get the most out of our data.

Telling stories is an important part of a product manager’s job that is often overlooked. If someone asks us, as a product manager, what makes our product different and better than the competition, we can give a long list of reasons to back up our success. It’s what it can do. It’s the price, how well it works, how new it is, and so on. And we’re probably telling potential customers the same things when we talk to them.

If we present our data as a bunch of unrelated charts and graphs, our audience might find it hard to understand or, even worse, come to the wrong conclusions. This can cause us to make bad decisions, which can hurt our business in a big way.

Here is a five-step framework that we can use to structure the storytelling exercise:

1. Identify the audience.

2. Develop a narrative.

3. Choose the right data and visualizations.

4. Draw attention to key information.

5. Engage our audience.

We'll now dive deeper into each of these aspects.

Identify the audience

Before we start telling our story, we should first think about who we’re going to tell it to. We should think about what makes them tick, what they’re interested in, and how to connect with them best. To win over an audience, we have to understand where they are coming from and what their priorities are, and then connect with them on an emotional and personal level.

When thinking about our audience, keep in mind that different people on our team have different goals and points of view. A good story based on data should talk about these differences. One good place to start is to think about how much our audience already knows about the subject we’re talking about.

Beginners may not know much about the subject but want to learn more. The bottom line is important to executives, and key statistics and KPIs help them decide what to do. We should pay attention to the most important lessons and how they affect the business as a whole. Business managers are mostly interested in themselves and want to make things better. We’ll have to show them how our ideas can lead to results that people can see and touch.

Analysts are interested in the details and want to know how we do things. We’ll have to show them how good our analysis is. Generalists care most about big ideas and big-picture analysis. Experts want information that is less based on stories and more based on questions. Supervisors want both details and insights that they can use. Executives are often in a hurry, so they want to know what the conclusions and implications are right away.

Develop a narrative

If we just show our data without explaining what it means, our audience will (at best!) look at it for a few seconds and move on, not really remembering any of the insights we’ve shared. A simple way of creating a narrative using different types of analysis is by answering the following questions in the following structure:

• What happened? Describe what happened using descriptive analytics. This paints a picture of the problem that we are trying to get the audience’s attention to, and this gives us a chance to emphasize the size of the impact our solution could have.

• Why did it happen? Once we have painted the picture of the problem, we should share insights into the cause of the problem that helped us identify the possible impact and solutions. The type of analysis that helps us diagnose a problem and understand its causes is called diagnostic analysis.

• What will happen? Use predictive analysis to show what happens if the problem is left unsolved. Answering these questions helps the audience understand the risk of the problem.

• How do we make good things happen? Having identified the problem and justified the size of the impact, we can now come to the proposed solution on how we propose to fix it. We can use prescriptive analysis to show how we evaluated different factors and determined the next steps.

All four types of analysis—descriptive, diagnostic, predictive, and prescriptive—help us create a complete narrative that walks our audience through the depth of our analysis.

Choose the right data and visualizations

Data visualization is not the same thing as storytelling. Visualizations are still a critical part of creating a compelling narrative. When done right, data visualizations help people compare and understand information and put stories in the right context.

How do we create great data visualizations? We do this by:

1. Picking the right data to show.

2. Choosing the best way to show our data.

3. Making sure our visualization shows what’s most important.

Find the parts of the data that show the exact points we want to make. Take out any information that isn’t necessary to our story. If we give too much information, it’s hard for readers to see the insights we want them to see. Use metrics and naming conventions that our audience will understand, such as "developer journey," "time-to-first-call," and so on.

Now that we know what to show, we can start creating the visualizations. Start by asking what we need to get out of the visualization, since each type has its own strengths and weaknesses. The following illustration shows us a simple way to choose chart types based on the nature of the information we are trying to present.