Power of Context in Data-Driven Storytelling

Author:Murphy | View: 29884 | Time: 2025-03-23 12:46:57

What is data-driven storytelling, and why is it so important?

Data-driven storytelling is a way of communicating numerical information through narration and visualization. Its goal is to engage the audience and help them better understand the main conclusions and trends.

Data visualization helps engage viewers. It's not enough to show raw facts or information. It's important to display data in a manner that grabs the viewer's attention and is pleasant to the eye.

The narrative is the story itself. The narrative is located at the heart of a data-based story, helping to create a coherent message that makes sense to the viewers.

*Context is a critical component in data-driven storytelling and, in my opinion, accounts for 80% of the story's success. The narrative and visuals comprise the remaining 20%.**

Today, we have abundant data, tools, and methods. However, even with all these resources, using them in practice is still quite challenging. Storytelling can bridge this gap. The decision-makers in today's companies, board members, directors, and managers, should not be concerned about data cleaning, analysis techniques, or specific tools. What matters to them is the interpretation, recommendations, and their potential implications. They want to be able to ask questions, provide feedback, and understand the answers. All of them will need data-driven storytelling!

But data storytelling will inevitably misfire if we don't recognize the context!

The context of a story has three fundamental dimensions: situational, functional, and data.

In essence, context refers to the background info that both sides of the story (i.e., storyteller and the audience) need to understand. It answers three main questions: who, what, and how?

Who?

It's about identifying who the story is for and understanding the relationship between the storyteller and the audience.

What?

It's about the story's main topic and what you want the audience to do or decide after hearing it.

How?

It's about picking the correct data and tools for analysis and deciding how to share the findings with the audience.

And what may happen if we don't do that homework right?

Well, a lot. Let me give you a few examples.

Defense of Czestochowa

Czestochowa is a southern Polish city, home to a famous basilica dating back to the 15th century. It has a revered image of Mary and is deeply sacred to many Poles. The city is also known for its 1655 victory against the powerful Swedish Army, a celebrated example of bravery and resistance.

When the Polish soccer team faces a formidable opponent and must defend their goal (which happens often), we Poles jokingly call it "defense of Czestochowa". We hope they'll triumph like the 17th-century defenders, but it's usually wishful thinking.

Unfortunately, that's also quite a common scenario for our stories when we don't recognize or care about the context. Most criticism regarding data (whether valid or not) stems from ambiguity or inconsistency. When data is unclear or contradictory, it's easier for people to question its validity. And, as it turns out, it is more difficult to defend our position.

Returning to my Czestochowa example, there was (only) one opponent. Let's imagine that there are more of them. Two, three, more, all of them! That's when we find ourselves trapped in

The crossfire of questions

Crossfire of questions. Source: image generated by the author in DALL-E 2

This barrage intensifies significantly when someone influential or authoritative raises doubts or concerns. Once that seed of doubt is planted, others might quickly follow suit, posing questions or amplifying the initial concerns.

The danger here is twofold. Firstly, these doubts and concerns can easily overshadow our story and supporting analysis. Secondly, when faced with a relentless volley of questions, especially if unprepared, there's a risk of appearing less credible or knowledgeable, even if the data and the story were initially solid and well-researched.

And we can contribute to this mayhem ourselves. Especially if, at some point, we may do something, unintentionally, of course, which will…

Implicate someone

That's a classic. And indeed, not a nice one.

Source: image by the author generated in DALL-E 2

When presenting data-driven insights, there's a risk of inadvertently putting someone from the audience in an uncomfortable position, especially if the data reveals sensitive or unexpected results. This can lead to tension and disrupt the overall flow of the discussion. Such discomfort might arise from data that challenges existing beliefs, implies unintended consequences, or highlights particular areas or individuals unexpectedly.

These situations can sidetrack the main data insights and derail the intended narrative. The individual might feel targeted or defensive, especially if the data seems to criticize their work or decisions. Furthermore, other audience members might tread cautiously, concerned that subsequent data points might spotlight their areas. It's crucial to approach data presentation with empathy and tact to maintain a constructive environment.

Ok. Now we know what can potentially go wrong. Let's look at how we can avoid or escape from difficult situations.

Situational context

The might of preparation

When crafting a data-driven presentation or communication, it's vital to set the proper context and understand our audience's dynamics, whether they are peers, superiors, or subordinates. Anticipating potential objections and ensuring clarity around the data used is critical. Pre-emptive data validation sessions with stakeholders can be invaluable, especially with unfamiliar tools or datasets. Decide whether a detailed presentation or a brief communication, like an email, is more appropriate. Set a clear goal for the communication, and always have a primary message ready for limited attention spans.

Consider stakeholder interviews, surveys, direct observations, and focus groups for a holistic understanding. Technical uncertainties? Consult field experts. Social media can also offer insights into public sentiment and potential audience perspectives.

Motivation is the key

For one of my presentations about storytelling, I crafted an extraordinary construction (in my subjective view). We all agree that stories in life appeal to emotions. Right? We identify with the hero, hate evil characters, and endure fear and joy. These are emotions. But how do you raise the emotions in the corporate life? Well, through motivation. Do you see the connection already?

And the motivation in corporate life has, effectively, two ends:

We make stakeholders happy or
We make them unhappy.

They are happy when they see good results, progress in the project, or the first positive effects of the strategy. They become unhappy if they see the opposite. Or they feel threatened by somebody else: an evil character.

In storytelling, three common adversaries shape the narrative. The first is competition. This external challenge forces businesses or individuals to innovate and stay ahead. Then, we have global uncertainties. The lingering concerns post-COVID and the effects of the 2022 war in Ukraine exemplify how such events can create widespread apprehension. These unpredictable situations often necessitate a change in plans or strategies. Lastly, there's the internal struggle. Issues like outdated technology might seem minor but can lead to more significant problems over time, making it harder to face external challenges. Together, these elements form the core obstacles in our story, setting the stage for overcoming adversity and growth.

Scenario analysis can be like a "what if" game for businesses navigating among the bad characters. For competition, it helps us guess our rivals' next moves and prepare. It helps us think about different outcomes and plan ahead for unexpected events like COVID or conflicts (at least to some extent). And for our issues, like old technology, it shows us the benefits of fixing them or the risks of ignoring them. It's a tool that helps businesses prepare for different situations.

Navigating out the challenging scenarios

The "defense of Czestochowa" trap emphasizes the need for preemptive action. Data must be presented transparently and coherently. Inconsistencies or ambiguous visualizations can compromise its integrity, leading to skepticism. Strong data management practices and precise visualization techniques rooted in Data Science principles are pivotal in reinforcing the data's credibility. It's essential to be meticulously prepared, anticipate potential challenges, and maintain a consistent data-driven narrative. A review by a data-savvy colleague can unearth potential pitfalls or biases in the analysis. When presenting potentially sensitive or controversial findings, attuning to the audience's responses is crucial. Swiftly address concerns, provide data-backed clarifications, and propose a follow-up discussion if a topic becomes contentious. In the dynamic landscape of data science, fostering an open, evidence-based communication environment ensures everyone feels acknowledged, and data-driven decisions are made.

Functional context

The functional aspect of context, especially when considering data and data science, is inherently more "grounded". This pragmatic dimension emphasizes the translation of raw data into actionable insights that resonate with stakeholders and have a direct impact on business decisions. While the broader perspectives provide a general overview, functional context ensures that data-driven narratives are not merely informative but are genuinely impactful, meaningful, and tightly aligned with business objectives. It's an element that cannot be overlooked.

Functional context is about making data useful and relevant for real-world decisions. It turns raw data into insights that can help a business. Situational context, on the other hand, is about setting the scene. It answers who should care about the data, what the main message is, and how it should be shared. Think of functional context as the "meat" of the data story, while situational context is the "setting" where the story happens. Both are important to make sure the data story is clear and impactful.

Data literacy

A critical aspect that helps to keep afloat the functional context meanders is data literacy.

Data literacy is like being able to read and understand the deeper story behind numbers. It's about really getting what the numbers are telling us and making sure we're sharing true and important information. This skill helps turn tricky data into simple stories that people can easily connect with.

When you're data-literate, you can spot trends, relationships or outliers and be sure that the data is of good quality. This makes the stories you tell with data both interesting and trustworthy. In short, it's about turning numbers into a clear story that everyone gets and believes in.

Relevance

Relevance is a cornerstone of compelling storytelling, especially if rooted in data science. It's about ensuring that the problem, concept, or opportunity you're presenting directly impacts the business. While it doesn't always need to be tangible, the more substantial the effect, the more influential your story will be.

When crafting your narrative, it's crucial to pinpoint the Key Performance Indicators (KPIs) and Key Result Indicators (KRIs) that resonate with your stakeholders. Understanding these isn't typically hard. Perhaps your organization uses a balanced scorecard, a tool highlighting how goals trickle down throughout the company, each with relevant metrics. Alternatively, being aware of the measures affecting someone's incentives, even without exact figures, can amplify your story's impact.

Grasping your company's value creation chain is vital. It bridges data literacy with relevance. For example, you can demonstrate to stakeholders how specific data-driven methods can enhance the efficiency of the customer journey. Take a look at the below diagram. Educating your audience about the benefits of specific data analysis tools in optimizing key business processes is a valuable investment we should prioritize.

Source: image by the author inspired by [1]

Plot points

The structure and flow of a story give it form, but the plot points make it unique. These are the key details that shape each scene. They allow the audience make their own assessment of the importance of presented data. They help them visualize the intended context. Here are nine essential plot points to consider:

Trend Shifts: This looks at whether a trend is rising or falling and how it's progressing. For example, even after investing in safety, the number of accidents on a production line might increase.
Dependency: Shows how two things are related. For instance, a higher Net Promoter Score (NPS) might correlate with more customer retention.
Intersection: This is about when one variable surpasses another. It could be positive, like when a startup's revenue exceeds its costs, or negative, like when a product's sales fall below a competitor's.
Forecast: Predicts the future. For example, how a country's population might change due to migration and other demographic shifts.
Comparison: Points out similarities or differences between two or more items. It could be comparing the efficiency of an old machine to a new one we're considering buying. This is often used in business stories.
Drill-down: Breaking down a general figure into detailed segments. You might see overall regional results on a dashboard, then delve deeper into sub-regions or individual stores.
Aggregation (Zoom Out): The opposite of drill-down. For instance, we compare one store's results to the regional or national average.
Cluster Analysis: Reveals concentrations or spreads within a data set. A significant cluster might indicate an opportunity or problem. For example, a clinic's most expensive patients might all live near specific factories.
Outliers: These are data points that stand out from the rest. An outlier can signify a problem or opportunity, depending on the context. For instance, a particular product might be bought way more frequently than others in its category [2].

Source: image by the author based on [2].

Data context

Data context refers to the additional information or "metadata" accompanying raw data to make it more understandable and valuable. This metadata essentially acts like a road map, offering crucial details such as who collected the data, when and where it was collected, and why it was gathered in the first place. Knowing the context allows for more accurate interpretation and effective decision-making.

Metadata with context :). Image by the author.

In a corporate environment, managing this metadata requires the cooperation of various roles:

Data Producers: The individuals or systems that create the data. They are responsible for ensuring that metadata is accurate and complete when data is generated or collected.
Data Consumers: End-users who rely on the data for various tasks, such as analysis, reporting, or decision-making. They need the metadata to interpret the data correctly and to trust its validity.
Metadata Managers: People or systems specifically designated to manage metadata. They make sure that metadata is stored in an organized manner and is easily accessible to data consumers.

The importance of proper metadata management cannot be overstated, especially in an increasingly data-driven world. Data can easily be misinterpreted or misused without this context, leading to incorrect conclusions and poor decision-making.

Technological advances have led to more dynamic methods of managing metadata, including "active metadata management". In this approach, metadata is not merely stored in a static state but is continually updated and synchronized across different systems. This enables more seamless integration and cross-referencing of data, providing a more comprehensive and up-to-date view. Active metadata management is crucial in environments where data is constantly updated, and quick, accurate interpretation is required for real-time decision-making [3].

Citing data source

One last thing worth mentioning concerning context is citing your data source. When you use data in your work, it's like borrowing a book from someone. Just as you'd thank someone for lending you a book, you should give credit to the people who provided the data. This shows respect and helps others trace back to the original data if they want to see it for themselves. It's a win-win: those who made the data get recognized, and those who use it can show their work is based on solid information. To give credit, mention who created the data, when it was made, and what it's called [4]. In my opinion, providing the proper credit can be a lifesaver, especially in those first three situations we talked about earlier. And if the data comes from someone well-known or a trusted place, it's even better.

Concluding remarks

Telling stories with data isn't just about numbers. It's about making those numbers easy to understand with visuals and a good story. The key is context: knowing the situation, making the data useful, and understanding where the data comes from. Think of it like explaining a famous battle: if you don't set the scene right, people won't get it. To tell a good story, you need to know your audience, pick a suitable theme, and be ready for any questions. It's also about connecting with people's feelings, especially in business. Being good with data means your stories are both exciting and believable. Knowing why data matters to a company and spotting interesting bits in the data makes the story richer. In a world full of data, the proper context is everything. Understanding where data comes from and why it's trustworthy is crucial for making good decisions.

*Author's subjective assessment

Did you like this article? Consider a subscription to get notified about new stories I publish or follow me.