Currently, every company dies in data. While a CRM has too much customer data, Marketing provides real-time metrics that can be viewed on a dashboard. While many transactions are happening at the same time in finance, and even while a supply chain sends out more telemetry than a person can track at any given moment, most organisations’ businesses seem unable to produce a truly data-driven decision.
The difficulty that organisations face is not due to a lack of data; it is that they cannot convert the data into knowledge, and in turn, convert knowledge into actions.
This explains data storytelling as the developing link between raw numbers and the final decision. In an increasingly commodified world of BI (business intelligence) and, therefore, the ability to turn data into a usable story, this skill is becoming one of the most sought-after qualifications within the modern organisation.
What Is Data Storytelling?
Data Storytelling is the construction of an account or story associated with data that conveys the key messages within that data to an identified audience in a compelling, memorable, and actionable manner. Data Storytelling connects three disciplines: data analysis, data visualization, and narrative communication.
Analytics provides the factual basis for your analysis, while visualization converts the facts into graphics; however, the art of Data Storytelling creates context around the facts through context, protagonist, conflict, and resolution. Analysis provides “what has happened” with facts and visuals, but Data Storytelling provides detailed context to what happened and its significance. In addition to providing a general context of what has happened, Data Storytelling provides context on why this matters. Who does this impact? And what are the next steps?
Data Foundation
Data is needed to produce accurate, clean & relevant reports; otherwise, the narrative would lack rigor and therefore have no validity, and all claims in the narrative must be supported by evidence-based documents included as an appendix.
Visualization
Visual representations of data (i.e., charts, maps, and graphics) provide a medium for conveying information, allowing immediate identification of patterns or relationships in the data and transforming abstract information into a more concrete form.
Narrative Arc
To assist the audience in understanding the data, the structure of the overall story (i.e., context, conflict, and resolution) allows the audience to derive meaning (i.e., create knowledge) from the data, rather than just being presented with it.
Audience Empathy
An effective data story is developed for the audience it will be presented to, and therefore, a CFO’s needs for the same data set would be far different from a Product Manager’s.
Â
Why It Matters More Than Ever
Business intelligence has undergone several changes over time. The first change enabled us to see long-term outcomes through structured reporting and dashboards that showed executives what had happened previously. The second change introduced self-service analytics, allowing non-technical users to perform their own analysis and explore data independently. The third change, which we are currently experiencing, enables insight activation by ensuring the right insights reach the appropriate audience in an actionable way.
The real challenge is a cognitive one: humans have very limited working memory. There are many studies indicating that people retain information much better when it is presented in a narrative format than when it is presented as data. For example, if I gave you a simple list of quarterly sales numbers, you would probably forget them quickly. However, if I presented the same information as part of a story about one region of the country that performed poorly and what actions were taken to turn the region around, targeting a new customer base, using an appropriate number of charts to support the story, you would remember the information much longer.
Â
Data Storytelling Across the Business
It’s important to note that data storytelling is not limited to one particular department or use. It’s applicable across all areas of a modern business.
Strategy and Executive Decision-Making
One way data storytelling is used at the executive level is to help CEOs understand what is going on in their companies when they have to make important decisions about how to operate them. Examples of this include CEOs making presentations to their boards of directors, providing updates to their investors, and participating in strategic planning. A CEO who presents a pivot strategy based on a well-told data narrative, such as the decline in margins on traditional products compared to the significant growth of a new product category, is much more likely to have the board (CFO) approve the decision than if they presented each number separately.
Â
Marketing and customer intelligence
Through storytelling, marketing teams can illustrate customer behavior clearly and usefully. Instead of simply providing attribution reports to stakeholders, experienced analysts build narratives that illustrate how a typical high-value customer journeys through the purchase funnel; at what point in the funnel customers drop off, and what campaigns have successfully re-engaged an inactive customer segment. These narratives provide a basis for marketing leadership to make budget allocation decisions, build more effective messaging, and justify investment in undervalued channels.
Â
Operations and supply chain
Data related to operations and supply chain can be very complex. They can consist of multiple levels of detail: supplier performance, inventory velocity, and demand forecast data all reside in the same data source. Data storytelling removes the complexity of data and presents it in a more accessible format. For example, an effective narrative for supply disruption events includes details on where the disruption originated, the downstream impacts, and modeling of possible outcomes of various responses. A well-constructed narrative for supply disruption events is much more effective than providing a spreadsheet of the underlying logistics data, reducing the time required to make the decision and providing a similar understanding to all cross-functional team members.
Â
Risk and Finance ManagementÂ
For hundreds of years, accounting has used numbers to tell stories and, with the advent of business intelligence (BI), finance departments have access to richer narratives about what is happening in their businesses. More than simply providing a numeric representation of historical events, BI (1) provides finance departments with new ways to assess the nudging effects of strategic decisions on their financial results via scenario models; (2) delivers financial reporting systems that not only measure but also explain why variances have occurred from a forecast; and (3) creates risk dashboards that provide executives who cannot quantify risks intuitively with an understanding of the range of risks present in their decisions’ outcomes, both statistically and numerically.
Data storytelling in finance builds trust in forecasts and models, ultimately generating confidence within an organization in its financial planning procedures.
Â
People Analytics and Human ResourcesÂ
People analytics is one of the fastest-growing uses of BI. However, there is often a level of sensitivity around the information being collected from people analytics (e.g., attrition rates, engagement ratings, compensation averages, etc.). “Data storytelling” within HR will require MAXIMUM care to present findings as opportunities rather than indictments, and to present stories with as much empathy as possible to build credibility for their organizations’ data-driven HR practices. A compelling story about the factors driving attrition among engineers (in this case) can create traction for HR’s retention-investment request, based on the sources of attrition, which could convert otherwise uninterested leadership to finance this investment.
Common Pitfalls to Avoid
Although very powerful, many data storytelling efforts are poorly executed; many organizations repeatedly encounter the same failure modes.
The Data Dump – One of the most common failures is data dumping. Data dumping occurs when someone presents all the data available rather than developing a focused narrative. More data does not mean greater clarity; in fact, it usually means less clarity. Good storytellers are ruthless editors; they select only the data that supports the core argument of their story.
Correlation without causation: Another common problem in data storytelling is using correlation between two events as evidence that one event causes the other to occur. Trust is lost when data scientists make claims of causality without the appropriate level of analysis; it is also lost when readers draw their own interpretations of causality from a lack of clear distinction between correlation and causation.
Audience mismatch – this is also a big issue. Technical audiences, for example, do not comprehend or respect the efforts of storytellers who oversimplify information. Alternatively, technical storytellers who use highly sophisticated language when telling stories to non-technical audiences will also lose credibility. One of the best characteristics of a really excellent storyteller is their ability to code-switch between technical depth and an accessible summary depending on who is in the audience at that moment.
Cherry-picking – One of the most detrimental pitfalls in data analysis is cherry-picking, the selection of data that only confirms a person’s already established conclusion. Cherry-picking is not only intellectually dishonest but also physically dangerous; most individuals will find that their cherry-picked narratives will fail once confronted with reality. For data stories to be trustworthy, information must include both evidence that supports one’s claims, as well as evidence contrary to those claims.
Â
Building a Data Storytelling Culture
To successfully embed data storytelling throughout an organization requires more than just employing competent analysts; it calls for changing the entire company’s approach to both how it produces and uses data.
Make investing in data literacy a company-wide priority
Data storytelling is only successful when your audience has an adequate understanding of the data to interpret it accurately and critically. To develop data literacy across the organization, you need to offer broad data literacy training to all employees, not only analysts.
Reward insight as well as reporting
If your analysts are evaluated only on how many reports they produce and how accurate those reports are, they lack motivation to develop their ability to craft stories from the data. You must recognize and reward the developing ability to translate data into actionable takeaways.
Always design for the audience
Every piece of communicated data should begin with a clearly articulable question regarding who the intended recipient of the communication is and what kind of reaction action you want them to take after reading your piece of communicated information. The two most important concepts in storytelling with data are the principle of always considering the audience.
Develop visual communication guidelines.
Use a standard set of organizational protocols that define chart types, colors, and how to label items. Following these standards will lower the mental effort required for processing information and produce a steadily increasing visual literacy level across the entire organization.
Provide feedback loops.
Provide a means for decision makers to provide feedback on which data stories yielded positive outcomes and which did not. The creation of these feedback loops will provide a complete connection between data narratives and outcomes, helping to sustain improvement.
Use AI, but do not give up on narrative.
The latest advances in AI can be used to identify unusual patterns, produce analytical summaries, and help create visualizations. These can be great sources of acceleration for data extraction. However, decision makers will still need to decide what information is significant, who the users will be, and why this connection is important. AI can help identify patterns, but data narratives will create meaning.
The Future of Data Storytelling in BI
While generative artificial intelligence (AI) is shifting the world of analytics, data storytelling has become more essential than ever before and will not disappear. With the advent of automated insight generation, there are now more stories on data available to humans than ever before; thus, the skills of selecting, shaping, and communicating the proper story to the right audience have grown exponentially in importance.
Moving forward, we need to see further advances in personalizing data stories. Business intelligence (BI) systems can generate different narratives (or stories) for different audiences, using the same underlying data source as an example. A store manager will require an operational focus, a regional vice president will require a portfolio view, and a CEO will need a strategic summary. The underlying technology is maturing rapidly, but significant human judgment will still be involved in structuring each of these audience-centric narrative layers.
Additionally, we are also beginning to see interactive and immersive data stories in a variety of formats, dashboards that not only guide users through the narrative but also allow for exploratory use; augmented reality visualizations that place data in real-life spaces; and conversational BI interfaces that enable end-users to interact with the data via natural language. These examples signify new formats of data storytelling that have their own unique “grammars” and conventions to be developed.
Â
Conclusion
Data storytelling is a hard skill at the heart of business intelligence, enabling BI to fulfil its core mission: helping organizations make better data-driven decisions that improve performance. In a world where cheap data is abundant and attention is a precious commodity, not being able to develop a clear, compelling, and honest narrative from that data is arguably the most critical component of the modern analytical stack.
Organizations that master data storytelling will not only analyze data more effectively but also enhance their decision-making, accelerate their actions, and develop cultures where decisions can truly be made on evidence. The true value of data storytelling to business intelligence has only begun to be realised.
