Storytelling in Analytics

5 Easy Steps to Make Your Data Tell a Story
To influence business decisions with analytics, sharpen your storytelling skills

So you’re a grand master of SQL, Python or R.

You also know ggplot2, NumPy and Pandas and you’re a wiz at data manipulation and visualization.

But is your analysis relevant to the business issue? And do your stakeholders buy in to your findings and recommendations? Can you articulate your thought process well enough to engage your audiences?

Maybe it’s time to pause your data analysis for a second and take a moment to think about storytelling.


With analytics, we are typically trying to influence decision making and drive changes. Whether it’s the head of marketing, a product manager or a program manager, those you need to persuade may not be as familiar with data analytics. But we are all comfortable with stories.

Compelling presentations and great storytelling skills are not natural gifts. They require a lot of practice.

In the UC Davis M.S. in Business Analytics (MSBA) program, we divide our time between conquering computing and quantitative problems and, in the other half, understanding how businesses work and on learning the basics of storytelling.


On a recent Friday afternoon, Darin LaFramboise spoke to our MSBA class and shared his storytelling experience. A UC Davis alumnus, LaFramboise is an analytics expert with more than 10 years in the industry. He was the analytics lead for the global business marketing team at Facebook and now is a business strategy manager with Atlassian. He has also spoken at UC Davis BAx Meetups.

LaFramboise shared how he engages audiences with fun stories in order to create a lively atmosphere. He adds humor and a clear structure to his stories (along with the right amount of data) to make the entire story interesting and informative.

Here are my five takeaways from LaFramboise’s talk:


At Facebook, LaFramboise said it typically takes six months to a year to hire a good analyst. A hiring manager first flags interesting resumes from the application pool. Then the HR department performs the initial phone screens. In the second round, candidates undergo a technical assessment, which usually involves an overnight exercise. After that, the hiring manager performs phone screens, then follow-up technical screens. The final onsite interview is a series of four or five back-to-back interviews of 30 minutes each.

The final round is all about communication. How well can the interviewee articulate himself? How long does the interviewee take to get to the point? A good analyst stands out with strong interpersonal and communication skills.


Before diving deep into a huge dataset, think clearly about the business issue at hand. What’s the problem you are trying to solve? What’s your ultimate goal? Who are your stakeholders and what do they care about?

Knowing your audience not only helps with your analysis, but also helps you anticipate questions.


When jumping into the data, make sure your analysis is independent from the storytelling. Don’t perform your analysis with the intent of writing interesting stories and don’t focus on the story until you finish the analysis. Instead, focus on finding patterns in the dataset. Why? Analysts often harbor opinions on data or business problems before actually diving into them. This unintended bias can lead to imprecise results and can be difficult to overcome.

It’s important to understand that it’s okay if you don’t find the patterns you expected to see. It’s not unusual to find nothing meaningful in the data.


You may find a lot of interesting stuff from your analysis. In fact, you may be excited that everything is actually important.


Now think about your audience: Do they care about every single point in your findings? The head of marketing may not be as interested in many of the analytical details in your 30-minute presentation. You may need to tweak those slides to focus on the business issues and your recommendations.

To avoid dead ends, try this five-slide presentation structure:

  1. Devote five minutes to each slide and another five for Q&A.
  2. Start with the problem statement, with a visual in your first slide.
  3. In the second and third slides, share the approach you took toward this problem and forecast how the business would be affected in the long term if you did not solve it.
  4. Next, lay out your hypothesis, and include your test data or resulting data.
  5. The last slide wraps up your presentation with the solution and the business opportunity it reveals.


Compelling presentations and great storytelling skills are not natural gifts. They require a lot of practice. LaFramboise rehearses in front of the mirror two or three times before ever going on stage and he feels more confident and more familiar with his story as a result.

Most of us worry about the Q&A sessions most, since anticipating those questions can be excruciating. The better you know your audience, the more prepared you are for the Q&A. Knowing what they care about means you can anticipate their questions and plan for supplemental data and information in your spare slides.

When I landed in the UC Davis MSBA program without any computing experience, I worried more about honing my technical skills. I gave little consideration to storytelling. LaFramboise’s fabulous presentation style and his unfettered ability to engage his audiences, however, made a huge impression on me.

I now plan to leverage every opportunity to practice my storytelling and presentation skills as I take this leap into the analytics industry.