In 1995, Kurt Vonnegut proposed something radical about storytelling: Every narrative follows predictable emotional arcs. “There is no reason why the simple shapes of stories can’t be fed into computers,” he said. “They are beautiful shapes.”
Many years later, research proved him right—data scientists analyzed over 1,700 stories and discovered six basic patterns that form the building blocks of all storytelling. These patterns, which include rises and falls in emotion and shifts in fortune, are mirrored in the fluctuations of data.
There’s value in extending Vonnegut’s thinking here. Just as a good story needs a compelling narrative arc, your business data needs to be structured and interpreted to reveal its underlying narrative.
When I say “business data,” I am referring to much more than anecdotes. After 25 years in the business, I’ve lost track of how many times I’ve heard something along the lines of, “Sure, we mine our data. I remember the four times that this [particular event] has happened.”
If you find yourself uttering something similar, you’re in good company. I’ve found that the longer the organization has been around—the more established it is—typically the less inclined they are to data mine.
Data mining is a systematic approach to examining vast datasets to uncover meaningful insights, including hidden patterns, unexpected correlations, and outliers that might not be immediately apparent. The insights gained through data mining enable organizations to make more informed strategic decisions and improve their predictive capabilities, especially when it comes to incentivizing customer purchases and assessing risk.
Data storytelling, then, is creating a narrative from your data—which you’ve extracted through mining. Storytelling helps frame those insights in a way that is meaningful—often in ways that are visual (through charts, graphs, dashboards) and also through narrative techniques to guide the audience through the data’s meaning.
Data storytelling is the future. According to Gartner, by 2025, data stories will be the most widespread way of consuming analytics, and 75% of stories will be automatically generated using augmented analytics techniques. But the data on those who are actively data mining and storytelling offer a different narrative: Approximately 75% of the data that companies collect remains unused, and 60% of investments in analytics capabilities go to waste because insights aren’t utilized properly.
These gaps between goal and actual use are a stark reminder of what is lost when we simply collect data without crafting a story—and a nod toward future opportunities.
Case Study: The Danger of “Perfect” Numbers
Years ago, I worked with a CPG company whose numbers—on the surface—told a perfect story.
We asked for an inventory analysis, and the warehouse manager shared that the company was within 99.5% of the $30 million of inventory the system said it had.
I was thrilled—that kind of number rarely happens. I asked to see a report so I could better understand their process, and it was only then I realized how messy things were. Line one was off by positive 72%. Line two was off by negative 209%. Line after line, the discrepancies continued. It was some kind of magic that overall average squared in the end.
A deeper dive revealed more trouble. It turned out that the products were in different warehouses. Items were mislabeled, and inventory was sitting in trailers outside that hadn’t been accepted. Shipments that had been pulled and were ready to ship were still sitting in the docks. By the end of six weeks, we were within 96% of inventory. But so much more was lost in this process: The company couldn’t fill orders. They received chargebacks from the retailers. Customers were rightfully frustrated because they weren’t getting what they had ordered. Products ended up getting discounted when they could have sold for regular price—had they been marked properly.
The truth is that you can make numbers say whatever you want. “What does it mean?” is a much harder question to answer—and a far more important one. This meaning emerges when we move beyond simply collecting data to actively engaging with it through systematic analysis and storytelling.
Data Storytelling
Having a knowledge of the analytics and how the numbers fit together is important, of course. But even with perfect accuracy—as our CPG case showed—numbers alone don’t tell the full story. The real value comes from understanding how to interpret and apply these insights to drive business outcomes.
A few months ago, an associate showed me some product data and asked what I thought of it. It was broken down by geography and showed a stunning difference in how the products were sold. The number one product in the East didn’t crack the top 5 products in the West.
Why was this happening?
We got the sales manager for the West and the sales manager for the East on a call and asked for them to drill down into the one product. We shared the data under consideration and started talking about why this might be. Turns out, the attributes that we thought were selling the product weren’t the drivers at all—it was the secondary attributes that were generating sales. The issue was that insight hadn’t been shared. Once that information made it over to the other side, we saw a fairly quick adjustment. Of course, there were still regional differences, but all of a sudden, the product went from not being in the top five to being number three.
Companies need to make sure they’re collecting the right information and asking the questions in such a way that the data is consistent, comparable, and easy to analyze. This way, they can get both the patterns from the mining and the learning from the narrative, enabling them to make the most informed choices for their customers.
The challenge, of course, is getting started. Despite the clear value of both data mining and storytelling, many organizations struggle to make the shift. It might feel like a big leap, but it’s one worth taking—especially when approached the right way.
Getting Started
Companies don’t need to start with a complete data overhaul. Whether you’re using QuickBooks or a multi-million dollar system, your data is there. The question is whether it’s accurate and meaningful enough to drive real insights—and if you’re the kind of company that prioritizes telling a story.
Here are some essential steps organizations can take to foster a data-centric environment:
1. Consider your culture.
“We’re doing fine” might be the most expensive phrase in business. Even when companies are successful, it’s crucial to maintain a mindset of continuous improvement. For small to mid-sized businesses especially, getting started with data can feel like a massive uphill climb, but it’s an investment that pays off. Don’t let self-doubt about your capability to change mask itself as contentment with the current state.
Remember, too, that sometimes the most effective changes start small, and a culture of improvement has ripple effects. Recently, I asked a junior team member to play devil’s advocate during an internal strategy session. “No matter what I say,” I told him, “take the opposite stance.”
At first, he struggled, but then he got into it. Did we trigger a wholesale change? Absolutely not. But we did move the needle 10% to the left or right—and sometimes, that makes all the difference.
A culture that values data storytelling is one where asking “What does this mean?” is as natural as asking “What are the numbers?”
2. Begin with Existing Data
Once you’ve begun to shift your culture toward improvement, the next step is to evaluate your current data landscape. To determine the quality of your data, consider whether you have one source of truth—one set of numbers widely recognized as “our numbers.” If finance has one set, sales another, and production a third, that’s a red flag. Ask yourself if you trust your numbers enough to make decisions based on them. The numbers simply have to be accurate: Ensure you’re looking at net numbers, not gross, that you’ve taken allowances into account, and that your standard costs are consistent.
This evaluation process requires careful attention to detail. You’ll need to trace data to its source, ensuring you’re looking at actual data, not manipulated information. This may mean reviewing invoices, purchase orders, and other primary documents. Then comes the essential but time-consuming process of building the data backup. Many organizations find that having an experienced partner in this process can make it more manageable and effective. The key is to find the approach that works best for your organization’s specific needs and resources.
3. Embrace AI as a Partner
With a solid foundation of reliable data in place, organizations can then look to modern tools to enhance their capabilities. Many organizations still approach AI with skepticism. But consider the lesson from Hidden Figures: When faced with the arrival of IBM computers, Dorothy Vaughan didn’t resist—she taught her entire department how to program. She understood that success meant combining human insight with technological capability.
AI isn’t here to replace analysis—it’s here to enhance it. Tools like Tableau and Yellowfin are changing the game, but they all require human input, evaluation, and delivery. The East vs. West Coast product case is the perfect illustration—technology helped identify the pattern, but it took human insight and collaboration to understand what was happening and devise a solution.
At the end of the day, the goal isn’t perfect data analysis from day one. The goal is to build an organization that increasingly makes better decisions based on compelling, data-supported stories.
Are you ready to tell yours?
To connect with Rob, please visit here.