Data storytelling is the craft of presenting data-driven insights in a compelling narrative form in order to drive action and change within an organisation.
By focusing on structured communication, data storytelling aims to solve one of the most pressing problems in modern business: how to translate data into business impact.
This piece is a broad overview of data storytelling, why it matters, how to do it, and what to expect when an enterprise suddenly starts taking business analytics seriously.
"If you want people to make the right decisions with data, you have to get in their head in a way they understand. Throughout human history, the way to do that has been with stories."
Storytelling is an old idea — even in business. Most companies are already familiar with the power of narratives. Stories form the foundation of all external communications, and stories are essential for internal communications as well. The value proposition is a story. The overall company mission is a story — and it better be a good one, too. Stories engage people like nothing else, and can bring people together to tackle the biggest projects.
A full decade into the Big Data revolution, we are beginning to see demand for a new kind of story: the data story. The thinking here is that a narrative form makes it easier to convey insights to a broader audience.
Companies are drowning in data, but despite significant investment in modern business intelligence, hardly anybody is getting any wiser. The digital revolution, promised by consultants and solution vendors, hasn't really happened yet. Many in business analytics believe that we need to rethink the approach. Data stories might be part of that answer.
Data storytelling, sharing the narrative behind the numbers, is a solution to the last mile problem of analytics. Today, the reality is that even the best data projects falter as the data hits the human interface. Failure to communicate analytics project results to the rest of the business defeats most initiatives. Nothing happens. Reports go unread, insights get ignored, dashboards fail to inspire the right action.
The race is now on to find the narratives that help companies make sense of all the data they have access to. We all need data stories to help us connect the dots and, ultimately, to change what we do and how we go about doing it.
This piece looks at data storytelling broadly in three parts. First, we'll take a closer look at data storytelling, why it matters, and what is on offer. Second, we'll consider the craft of storytelling, with a focus on some of the particulars of data storytelling. Third, we'll consider the audience: colleagues, managers, and the rest of the organisation.
Along the way we'll draw heavily on popular business press, management consultants, and market news.
The Language of Insights
The World on Data
Software ate the world and almost overnight every single industry was transformed by data.
From the beginning, companies had high hopes for their data initiatives. Data engineers and data scientists, hired at great expense, were left alone to figure out the details. These specialists put their heads together and built the most elaborate pipelines they could. The data natives figured out how to do processing and storage at scale, and to do it all up in the clouds. Things were happening. Surely it was just a matter of time.
Months and years rolled by, but the promised data-driven world never really arrived. Desperate to outplay the competition, companies kept on investing in new data people and new technology. But project after project, these initiatives struggled to deliver transformative change.
Analytics projects often start strong, but quickly lose momentum and eventually lose their way somewhere in the final mile. And so, today, decision makers everywhere continue to make do without their data-driven insights.
Data storytelling seeks to reinvent our relationship with data by focusing on the way we communicate data-driven insights. MIT Sloan designated data storytelling as one of the breakthrough ideas of 2020, with this definition:
Interpretation of data and information in a narrative format that helps decision-makers absorb insights and initiate appropriate actions.
Data storytelling is founded on an ancient idea — stories and their telling — but it's very much concerned with a modern problem, the data deluge. It's highly technical and quantitative on one hand, but focused on people, communication and all things qualitative on the other. It's inherently social, but in a way builds on private scholarship and study. On one hand it's all business, but on the other it's fundamentally creative. It requires both technical mastery and a sense of humour and style.
Data storytelling is the perfect craft for the digital age, because it bridges the old world and the new.
Brent Dykes (@analyticshero) believes people, not just machines, will be key in helping us extract value from data. With a basic level of data literacy and general know-how around data tools, individuals across the organisations can apply their domain expertise and uncover meaningful data insights. Sadly, many of these insights never amount to anything. They get ignored or squandered simply due to poor communication.
Dashboards, other interactive workflows, and automated alerts are useful for precise, role-specific insights, but these approaches fall short when scaling things up. As the potential impact of data-driven insights grows, analysts need to engage with others. They may need to win buy-in from their colleagues and managers in order to secure the necessary resources and support. Inspiring action and change within an organisation requires effective insight communication. This is where data stories come in.
"While access to data is improving," Dykes writes, "the communication of insights still lags behind where it needs to be — even for the analytics and data science professionals who have been typically tasked with this responsibility." The challenge is crafting a compelling data story.
"Facts and figures alone won’t influence decisions or move people to act, but well-crafted data stories can and will," Dykes argues. Data stories, structured narratives that communicate data-driven insights, often rely on various kinds of visualisations and other presentation, but in a sense the exact methods are not the point. Skilled data storytellers focus on the human story hidden within the data, and emphasise creativity, empathy and contextual understanding. Dykes believes this cannot be easily replicated by technology.
Data storytellers can play a central role in unlocking the value surfaced by data initiatives. Storytellers have an important communication role in the last mile, helping people discover and understand key insights. Stories can help people make better decisions and take decisive action.
There's a growing need for skilled data storytellers, but really every role has become increasingly data intensive. The whole workforce should be "data literate" to some degree. The greater the change, the more people have to be ready to engage with data. "Your digital transformation efforts cannot tell their own tales," Dykes writes. Data storytellers are the vanguard and will "play an integral role in defining the next decade of data".
Why Data Storytelling
Cassidy Shield from Narrative Science, an augmented analytics vendor, interviewed Dykes for their podcast series. They discuss Dykes' book on data storytelling and cover a lot of ground. It's a great introduction to the subject.
Dykes formulates data storytelling as a process taking an insight and communicating it clearly and effectively to another group of people. Data storytelling is more than just visualisation, the narrative structure really is key. With a story there's a direction and the listener-viewer can appreciate the journey. If there's no climax or call to action, it's more difficult to have an impact.
Stories are important, because they get through to people on the emotional side. As an analyst, the common mistake is to think that just showing up with the data and the reasoning is enough. "Emotion is a big part of decision making," Dykes asserts. Stories help people connect with insights. It's not enough to just dump data on people. The insight has to be there, the heart of the data story.
Building a data story begins with an insight about data. Sometimes the customer can provide the observation; maybe something looks unusual. With the insight in hand, the next step is to figure out what minimum context is needed to tell the story. You don't give all of the context, just enough to build up to an "Aha!" moment. Five whys helps here.
Refining the story is about defining the problem and the solution. Remove the noise, all the unnecessary detail. Data stories and dashboards, for example, serve a different function. Dashboards support overviews and "drilling down" workflows, focusing more on exploration and framing. Data stories shine when the insight is more subtle or context makes all the difference. A data story helps people connect the dots they see perfectly well in isolation.
When it comes to interactions with executives, Dykes argues that a "just give me the facts" mindset is in fact a coping mechanism. Business is hugely complicated, and executives generally don't have the time to spend on sitting through long form narratives. Dykes suggests using "data trailers", brief invitations to hear the full story, if there's interest. "Here's this problem, would you like to hear more?" It's useful to have an abbreviated version, 5-10 minutes, at hand.
For Dykes, analytics tech is there to support story building. Tools can help us find insights faster. Augmentation takes this to the next level. Tech can serve as a guide to the whole organisation, becoming a launch pad or a starting point for discussions all over. Over time, businesses can become more data literate and able to identify insights everywhere. A data literate company is itself more nimble and able to adapt and change.
In a later stage, one can start talking about data curiosity and data democratisation — both founded on open access to data and easy-to-use tools. Lead storytellers can have a great impact on company culture, but data stories really are a shared responsibility in organisations.
Data stories can be the currency of company culture.
Dykes asserts that domain expertise is the Achilles' heel for all analysts. Data storytellers, analysts in general, should work closely with business to make sure that their work is aligned with the business. Dykes sees the analytics team serving in a kind of personal trainer or a coach role in the business. Data storytellers help people tell their stories.
"Skilled data storytellers can magnify the analytics impact," Dykes posits. Lots of analytics projects fail in the last mile, as the analysis shifts from tech to people. This is often due to poor communication one way or another. At the same time there's more to the last mile than just storytelling. For example, UI and tooling matter a great deal.
On visualisation, Dykes points out that there once was a time when statisticians were not big on visualisation. But then some researchers showed how the same descriptive statistics can result from very different data sets, and people started thinking about data visualisation more. At the heart of all visualisation is a desire to make patterns, trends, and anomalies visible. Seeing dominates our perception.
For getting started with data storytelling, Dykes puts the emphasis on business analysts first knowing their audience. The stories will follow. Spending time with key stakeholders eventually yields the right insights. Analytics teams should not be walled gardens, but rather be forward deployed into functional teams. Analysts are always amazed by what business people do.
As for CEOs, Dykes recommends that they lead by example. CEOs should request action-driving data stories and invite analysts to share insights about the business. Company culture is about creating an environment, and CEOs can, if they so choose, breathe new life into data initiatives and the way data is used in the company.
👉 For more Brent Dykes, check out his book Effective Data Storytelling: How to Drive Change with Data, Narrative, and Visuals. (2019, Wiley)
Context and Empathy
"It’s hard for a dashboard to explain why something is happening," reports Beth Stackpole for MIT Sloan. Charts, dashboards, and visualisations sometimes fall flat with their intended audience. Humans need context and connection — a narrative — to comprehend what it all means. There's more to data than what any chart can capture.
Analytics dashboards can alert business users that some change has occurred, but dashboards cannot give the full story. The insight is in the context: a story helps make the change more intelligible and easier to communicate.
"Sometimes it’s a matter of overwhelming recipients with too much data; other times, it’s about presenting the wrong data", Stackpole explains. Without an effective narrative, insights fail to resonate with recipients.
Great storytellers anticipate the audience reaction, and same is true of data stories. Data storytellers need to be wise to how an audience is likely to respond to the analysis. The message should always be tailored to a specific audience, to help people take away the right insights and to initiate action. The key thing to remember is that not everyone is equally fluent in making sense of data — but there's a story for everyone.
Stackpole suggests that data scientists typically have advanced skills in the technical sphere, but rarely are able to articulate why they are doing what they’re doing. In a sense the whole data science pipeline is wrong. Data scientists have a hard time working backwards from questions into practical business solutions — they can't talk business.
Effective data storytelling is about focus. Remove the noise and make sure the insights grab people's attention. Graphical presentation of data is a key technique, but knowing what to show is the main thing. Effective communication is about getting the point across in the most direct, succinct manner.
"Perhaps the most difficult data storytelling skill to master is empathy", says professor Miro Kazakoff of MIT. Understanding where the audience is coming from, what kind of data analysis they’ll react to, is key to engaging data stories. Storytellers must be able to interpret different viewpoints and to present relevant material.
Data storytelling is clearly a key skill for the Big Data era. Enterprises are anxious to turn their vast data reserves into actionable business insights. But without context and empathy in the presentation, insights will not get leveraged in decision making.
Dykes, also interviewed, argues that data literacy is a key skill not just for data storytellers, but the broader workforce as well. A successful data-driven culture requires that both the presenter and the audience speak the same language. You need everyone to pull together to get organisations across the last mile of analytics.
Kazakoff agrees: "Being literate with data and able to explain the stories it reveals is as important a form of literacy as being able to read, write, and speak clearly. It’s a core skill, not a job function, and it cuts across all division and roles at a company." All jobs can be informed by data. There's no escaping the fact that everybody needs to be able to understand data and to explain to others what it means.
The history of business analytics can be seen as an exercise in the automation of insight generation.
In the beginning there is always just a pile of raw data. When we organise and assemble what we have into standard and ad hoc reports, we transform data into information. We build tools and analytical systems to support this work. As the industry moved from OLAP structures to descriptive modelling and further to predictive modelling, we have built ever more elaborate things, hoping to push from information to more easily absorbed knowledge.
Today, with AI and prescriptive analytics, we are pushing further still, supposedly towards wisdom and action. In many ways it still remains to be seen if we can build tools to help us in these higher tiers. The struggles of the last mile, and the biases that we often bake into our analysis, suggest that there might be more to wisdom than mere processing. Should we be more careful with this "unearned wisdom" that isn't grounded in personal experience?
Some believe that advances in technology will be enough to "bring clarity to a world of ambiguity", as Gartner themed its Data and Analytics Summit in 2019. Many business intelligence vendors are looking at data storytelling, and many believe that augmented analytics, AI powered analytics automation, is the answer.
Cassidy Shield of Narrative Science was at the Gartner summit and found broad support for their augmentation route, using AI to generate reports, stories and knowledge. Certainly the direction of travel is to move beyond dashboards.
Stu Kendall, also from Narrative Science, believes that generating natural language is the way to go for data stories. Kendall quotes Gartner research that predicts that by 2025, 75% of stories will be automatically generated. Narrative Science see augmented analytics as an assistive technology, helping the human analyst with data prep and insight crafting. Modern BI tools augment the analyst, and upgrade the workflows around data exploration and analysis.
Kendall writes that the Narrative Science team is excited for the future of business analytics. With data storytelling and augmented tools users, even non-specialists, can craft better stories faster. The juxtaposition with previous generations of business analytics is striking. "The past was marked by too few organizations having the skills to effectively scale data storytelling to everyone in the company."
Data storytelling can help decision makers engage with data and analytics, Kendall argues. In the past insights were packaged into dashboard and reports, but data storytelling is all about new, compelling forms of insight presentation. These new forms are more readily consumed and assimilated than those that came before.
And so the business analytics race is on. The challenge for vendors everywhere is the automation of data-driven story generation. There should be plenty of room for new thinking in this space as well.
"Storytelling became an umbrella term for a more humanized approach to convey information, especially for the digital era."
Stories are the language with which we share our insights and understanding. Both about current events and the bigger picture. Stories are inherently entertaining. We can't help engaging with our imagination, when somebody begins a piece of fiction. In the same way, in a business context, we are curious about what insights we might gain when somebody begins a presentation. Every story is an opportunity to plant seeds of understanding and action.
"Data storytelling is changing the way we communicate data insights – for the better," writes John Ostrowski for Ladder, a growth marketing agency. By shaping facts for easier consumption, a data storyteller can reach a broad audience and have an impact in an organisation. McKinsey analysts Bisson et al. agree: organisations have lots to gain from making analytics extremely user-friendly and tailored to each group of decision makers.
Storytelling is a form of translation. Data storytellers take quantitative data-driven insights and transform them into something more easily digested. It's all about finding the right language.
The narrative form helps people absorb information beyond what visualisation can do alone. The story gives context and colour to the facts, and helps the recipient make sense of new information. The story supports the insight and activates the listener. And compelling insights, backed by data, inspire action.
Data storytelling goes a step beyond dashboards. It’s hard for a dashboard to explain why something is happening, but a data story can help the audience connect the dots. A data story provides extra information or a view of the big picture — some relevant background information that gives raw and processed data a deeper meaning.
In the same way that a visualisation reveals patterns in data, a data story brings out the meaning in information. Meaning leads to understanding and perhaps even wisdom eventually. At the same time no amount of business intelligence alone will make things happen in an organisation. The last mile is a people problem.
Or, to frame it more proverbially, analytics results can show the door, but the organisation has to choose to walk through it. Data storytellers are the principal guides of this transition.
Stories are about communication. Aggregations and different visualisations help with communication, but facts and figures alone won’t influence decisions or move people to act. It takes a carefully crafted story to really get through to people. And it's always personal.
To have an impact on company culture, to drive transformative change, you may have to win over everybody or at least a large majority. Sometimes, for controversial or foundational change, you have to proceed one person at a time. To do this, you need to have a compelling data story both for yourself and the people around you. Data storytelling is so much more than visualisation: it's about taking insights across the last mile to drive actions for end users.
The Mechanics of Data Storytelling
Elements of Storytelling
Eric Avidon, writing for TechTarget, outlines the data story crafting process: "Data storytelling essentially consists of an analyst orienting themselves with a set of data, applying visualization techniques, reaching a moment of insight by creating a narrative around what they've discovered in the data and sharing that narrative to inform a decision."
In other words, data storytelling can be seen as a two stage process: the analyst first looks for insights and then communicates them in a narrative form. The challenge here is that these two tasks require entirely different skills. Analysis is intense, often systematic private work, while engaging communication requires the exact opposite approach. To excel at both is a rare thing.
Data storytelling is genuine storytelling, in the sense that many techniques from engaging fiction can be used with data stories as well. For example, the setting matters a great deal in fiction, and similarly context is essential for a great data story.
In fiction, the setting is best established indirectly through the experiences of individual characters. Similarly a data story can establish context by focusing on a single user, customer or stakeholder — real or hypothetical — and their experience with a product or a service.
In either case, heavy-handed exposition and world-building makes people passive — inside and outside the story. Similarly, a carefully chosen visual summary can give a rich impression of when and where the story takes place.
Data storytelling is fundamentally an exercise is moving an insight from one brain to another. To this end, pretty much anything goes; there are no fixed rules. Storytelling can be as multi-modal as our senses allow. We are dominantly visual by nature, but for stories in particular we are highly aural as well. In some situations we can be immensely tactile and spatially aware. Occasionally even our weaker senses can prove highly discerning.
The storytelling of the future could well make multifaceted use of a whole range of human capabilities.
Our senses reveal the world to us in layers, and we perceive the world at multiple levels simultaneously. We can think about the big picture, but we can focus on the details as well. Great fiction often presents a larger story through smaller, more personal narratives. We connect with personal stories, and with each one we hear, we gain a deeper understanding — a deeper appreciation — of life as a whole.
Empathy provides the ultimate context. Empathy is the most challenging part of storytelling and the most difficult skill to master. Great data storytellers understand where an audience is coming from and which parts of the data story they will react to.
How to Tell a Story
The best stories are memorable, impactful, and worth retelling again and again. Data stories are not that different. Data stories can look to the past or look to the future, but they are always grounded in the present. There is usually a call to action, and that almost always comes with an expiration date.
"At the end of the day, what differentiates stories that stick from stories that are lost, is the narrative’s engagement," writes John Ostrowski for Ladder.io. Ostrowski has assembled a collection of practical advice on data storytelling.
The main thing is to keep things simple. The purpose of data storytelling is to communicate the output of a data analysis exercise. The idea is to translate raw analysis results into clear narratives that can influence business decisions. Conciseness is key. The storyteller should filter out unnecessary details, map the content into something the audience can understand, and summarise the main points in some appropriate way. Cut the noise.
- WHO: Who receives the information and how will it be interpreted? Will the recipients trust your analysis — and you as the storyteller?
- WHAT: What insight are you trying to communicate? What action do you want to drive? Why does it matter?
- HOW: What is the best way to communicate the data story effectively? What could help the story "stick"?
"There's no point getting extra technical with stakeholders, if they are not concerned with the method", Ostrowski writes. When communicating the finer points of your analysis, context is everything. Details matter, but the things that really drive business decisions typically operate at a different level.
Ostrowski suggests that managing attention is the key activity while presenting. Effective storytellers bring attention to points that support the insight they wish to deliver. "Reduce cognitive load. Always use visuals that are familiar to the audience." Cognitive load is a mental barrier between your audience and your insight. Avoid confusing material, and help people understand your message without asking too much from them. Effort leads to disengagement.
A discussion on effective data visualisation is beyond the scope of this piece, but Ostrowski has some good tips there as well. The main thing is to maximise the amount of information that can be understood at a glance. All presentation should be aligned with how humans perceive things, with our cognitive abilities and limitations. A great rule of thumb is that every visualisation should answer a particular question — and do so at a glance.
"The folks who are doing the most interesting work on the storytelling front are journalists," say Haroz, Kosara, and Franconeri from Kellogg School at Northwestern University. The claim is that journalists, trained storytellers, have best understood the lesson that data presentation depends on the story being told. Visualisation makes a story memorable and supports persuasive communication.
Franconeri and his colleagues point out that the field of visualisations is still fairly young, certainly for complex data sets. The rules of visual data presentation are not really taught as such. "You take writing classes in college. You don’t take a graphical communication class." Not yet, at least.
The researchers have started studying the techniques and styles of visualisation, trying to figure out what works best and when and why. They frame data presentation as a kind of a guided tour. Franconeri is excited about innovations such as the connected scatterplot.
Visualisation, data presentation, is all about helping people understand the big picture. When it comes to convincing the audience, visualisation is not optional — it's necessary. Effective visualisation can have "a multiplicative impact on how well you can convey your ideas to people and how well those ideas will actually sink in and then lead to action".
Meredith Somers, reporting for MIT Sloan, writes that successful data storytellers recognise what is and isn't important. They present information without injecting bias and they keep things simple. Great stories are not accidental strokes of luck, but are rather shaped by a process of ruthless editing. Storytellers avoid adjusting data to fit a predetermined story, and make sure that insights are framed in a way the audience can understand.
The most basic story form is the three part narrative. Every story needs a beginning, a middle, and an end. Data stories use these three parts to communicate an actionable insight. Sometimes there is no action to take, in which case the data story serves a different function as a prompt for a conversation — ideally about an action plan.
"Create a conversation, not a presentation," writes John Coleman for HBR. His thesis is that structuring presentations around "the great unveil" — sharing key findings towards the end — is actually counterproductive.
Coleman's reasoning is that "one-sided expositions" lead to ineffective conversations, as people don't have time to process the news before the wrap-up discussion. Unveiling also makes exploring the problem space more difficult: it's hard to discuss alternatives in the middle of a narrative build-up. Discussion shouldn't be seen as a distraction either by the presenter or the audience.
In fact, fostering conversation is kind of the point. Discussion helps people understand the insight being shared. Coleman suggests sharing content before the meeting, and explicitly inviting discussion with sample call-out questions in the material. Presentations should have enough detail to be read and understood in advance. People should be informed going in.
During the presentation, it's a good idea to start with an executive summary that lists key conclusions. Another thing to try is facilitators, pre-appointed individuals — meeting chairs — who draw out comments and questions from the whole group. Find ways to activate people, to keep them engaged and participating.
"Communication between groups of people is most effective when participants are engaged, and the discussion is both inclusive and collaborative. Creating an ethos of conversation, rather than a one-sided presentation, for critical discussions can better leverage the collective intelligence of the team, make solutions to organizational problems better and more comprehensive, and improve ownership for execution of ideas."
John Bershin, writing for HBR, argues that good presentations make people uncomfortable. According to Bershin, presentations have a bad reputation because they are often badly made. A good one takes hours to build: research, clear thinking, and great care towards word choice, visuals, and flow. But great presentations are worth it.
"Presentations can help us do something more effectively than almost any other communication tool at our disposal," Bershin argues. "They enable us to make a compelling, persuasive argument." A great presentation gets through to people without overwhelming and without a fire hose of information.
If you want a group of people to adopt your point of view, as a storyteller, you should start with some difficult or painful issue people in the audience have. Set up the problem, suggest a solution, provide an action plan, and describe how they’ll be better off as a result. Sell your idea to the audience.
Nolan Browne, also writing for HBR, believes that the linear structure of a PowerPoint session is fundamentally flawed. Slideshows discourage interactivity, are terrible for information sharing, and limit audience participation options. In short, slide sets can easily obfuscate rather than clarify. "The best presenters tend to show rather than tell, creating opportunities to engage and persuade."
For Browne, presentations should be more like conversations. Presenters should solicit feedback and help listeners feel ownership over the ideas under discussion. This inspires thinking and action, and helps build bonds between people. The ideal presentation tool has "the necessary elements to support questions and intellectual digressions", but built-in flexibility to support varying amounts of data. Presentation tools should help with effective insight communication, but discourage "Chartjunk". There is a market for new, non-linear presentation tools.
Proficiency with data storytelling means recognising what is and isn't important. To a significant extent, data storytelling is about removing noise and focusing people’s attention on the key insights. Great data stories keep things simple. For all the editing work that goes into preparing data stories, it's important to avoid injecting bias into the analysis or the presentation.
A Data Story Playbook
The playbook is organised around five steps:
- Define the purpose of your story
- Conduct an analysis of the audience frame of mind
- Focus on structuring your story effectively: consider the three-act play outline
- Rehearse your story with colleagues; take and give feedback
- Always have a "speedthrough" version ready
The speedthrough version is a 30 second version of your story for settings where you have an unexpected opportunity to make an impact, such as a chance meeting with an executive in a hallway.
Data storytellers need some business expertise, and skills with data, analytics and tech. However, Little emphasises the role that soft skills play in data storytelling. Impact is built on an ability to communicate and a willingness to serve and to lead. Effective storytelling is about persuasion and translation. The successful analyst has a customer orientation, creativity, initiative, and resilience.
Little elaborates in a Forbes piece on why storytelling is such an important tool for a business analyst. In her view, it's the main mechanism for activating the organisation. As an analyst, storytelling is your best bet for influencing the people who determine where budget funds get allocated.
The challenge of course is that your recommendations may disturb the status quo. With a data story, you are asking people to change what they’re doing: to stop doing something they do now, to start with something they are currently not doing now, or to change behaviour in some way. Convincing people that this is the right thing to do is not easy. Show your stakeholders your data, your insight, and your recommendation.
Shaping a Story
Scott Berinato shares his three simple steps for telling stories with data in a rapid fire HBR Quick Study. For Berinato, a good data story has three parts: setup, conflict and resolution.
The setup is some reality, the state of the world before the change. It might be real, or one created just for the story. Conflict represents the change in this reality. "Without change there is no story." The key question to answer is: What is causing this change? Finally, resolution explains what the new reality is, the state after the change.
For Berinato, shaping the story starts with the data analysis, the process of finding these three story elements in the data. The main thing is to focus on the most important information. Berinato suggests distinguishing each stage with a separate image or a descriptive title. Highlight the important information and leave everything else out.
"Narrative is the most powerful, most human tool, we have to communicate. If you can apply storytelling to your data, it creates an emotional connection with the audience.
They're not only going to believe what you show them, they're going to feel it."
In his sharp TEDx talk "Seven Keys to Good Storytelling", Josh Campbell shares his tips for telling engaging stories. Campbell's view is that good stories are good because they avoid the things that make stories bad. While intended for general storytelling, the tips apply nicely to data stories as well.
- Be prepared, but not too prepared. Have an outline, stick to it, but have escape routes.
- You don’t have to be funny, because you are not.
- Name names. This is your story, you have all the power.
- Make peace with your story before you get on the stage. Storytelling to an audience is not a therapy session.
- Start from the beginning and end at the end. Trust your audience that they will get the lesson that you want them to get, or they will get the lesson they want to get. Don't rob them of that option.
- The devil is in the details, and the details are boring. Trust your audience. You are sharing human experiences and they will connect with you so skip the boring details.
- The audience is on your side, keep them there. Don't be controversial or offensive unless you have to.
Ultimately, Campbell strikes a slightly different note. Storytelling isn't as formulaic as one might hope. Sometimes details are important. Sometimes morals are necessary. Sometimes you have to offend your audience to make them listen. Practice telling your story.
Ben Wellington - Making data mean more through storytelling: How to become a data storyteller. The main thing is to connect with people's experiences. There's lots of data out there: focus on one idea, convey one idea. Aim for impact, keep it simple. "You can tell so much about a city, just by looking at the data."
Frank Evans - Moving Beyond Data Visualization: Dashboards, considered. Building data applications and explorable worlds. "The purpose of the car is to go somewhere. You don't drive a car with your eyes affixed on the dashboard, but on [the world behind the windscreen]."
Cole Nussbaumer Knaflic - How to Declutter Your Data Visualizations: Decluttering visualisations, an introduction to visual perception and visual design. Leverage how people see, employ visual order, create clear contrast, don't over-complicate, and strip down to build up.
David JP Phillips - The Magical Science of Storytelling: The cocktails you serve to your audience with your story. Make use of suspense, empathy, laughter; don't irritate your audience, help them de-stress. Functional storytelling: everybody is a storyteller, write down your stories, index those stories.
The Human Interface
"Analytics platforms largely focus on every aspect of the analytics process leading up to interpretation. They're about preparing the data for analysis rather than the analysis itself."
"The self-service and visual paradigm that now dominates BI is a limiting factor," writes Eric Avidon in a TechTarget report. Self-service is only as good as the individual's ability to serve themselves, Avidon argues. Gartner estimates that only about 30% of employees in most organisations use analytics as part of their job. Despite advances in augmented intelligence and ease of use, many people are still missing out on data-driven insights.
Data storytelling is set to change that. Data stories help people share insights in a more accessible way. The challenge is to find the best ways in which insights can be pulled out of the computer.
Self-service analytics will not go away entirely, but business analytics will be disrupted by augmented intelligence and data stories. Narratives alone will not be enough either. If data story curation is left to data scientists and analysts, some kind of a data literate specialist tribe, then the company will not be transformed by data. Everybody needs data literacy and the skills to interpret and develop narratives for use in data-driven decision making.
"The No. 1 metric of success of analytics with ordinary users," says James Richardson of Gartner, "is sustained adoption of the analytic technologies that are put in front of them. If you think about data stories as a trigger to adoption, you can very quickly see why this matters."
We are headed for a world where companies may employ some analysts and perhaps some self-service, but for most business analytics, machines will be doing the heavy lifting.
Avidon continues in another TechTarget piece, asserting that data-driven storytelling has the potential to bring about a revolution in analytics: "Data-driven storytelling tools will always be about extending the reach of analytics to a broader audience, and for the first time, potentially everyone." Moving beyond dashboards and reports to compelling narratives can change company culture. Data storytelling helps the whole workforce overcome the last mile problem.
"Visualizations helped make data more digestible and augmented intelligence is making analytics easier for untrained users to navigate," Avidon writes. In the same way, data-driven storytelling can expand the reach of data-driven insights. Making things more accessible has always been an innovation driver in this space.
Avidon points out that, informally, stories already permeate organisations. It's just that the narratives people use today are mostly unstructured. Data-driven storytelling gives information context and purpose, and makes insights more memorable and understandable. Data stories help people make better decisions. Part of data storytelling is understanding when a story needs to be told, when there's a story that needs to be shared.
Speculating about the future, Avidon suggests that storytelling platforms could try to break free from the simplicity of linear structure. "The real world is far more meandering."
Further, Avidon points out that data-driven storytelling platforms don't really know their users yet. This limits the kinds of stories that can be sketched automatically. The future promises personalised narratives, once the machine learning mechanisms reach that level of sophistication.
There's also an industry wide push towards prescriptive analytics or proactive analytics, i.e., having the analytics systems offer instructions. The ultimate data storytelling system would be hyper-personalised, telling the user not only what happened and the trends behind it, but also what to look out for and what could happen in the future.
Avidon predicts that the ascent of data storytelling will bring the fall of self-service analytics. When data stories are generated by automated means, there will no longer be a need to lean on the select few who can do the analysis by hand. At the same time, machines already best humans at spotting patterns and anomalies in data.
Business analytics veteran John Hagerty writes in agreement in a 2019 post. Hagerty is critical of the self-service model, suggesting that the paradigm can be a limiting factor for many users. Simple visualisations work fine for some data sets, but with complex data sets and "intricate insights", you need a proper story.
Hagerty argues that everybody reads graphics in their own way. There's lots of room for unintended interpretations. Business data can be easily misconstrued based on subjective views. In other words, dashboards will never be enough, they'll never bring people together.
If the goal is to use data to activate everyone in the company, to get everyone aligned, self-service isn't the solution. Data-driven insights need clear narratives. Hagerty hopes that augmented analytics and automatically generated data stories have something new to offer in this space.
Simply put, self-service won't be good enough in the future.
"The story is often more important than the data itself. Make sure you share insights in a story to build better understanding and drive the right outcomes for your organization."
The history of business analytics is full of tech innovations, but there's another important trend just below the surface. As business analytics has evolved, more and more people have gained access to data. And with data, people can more easily argue for doing things differently. This data democratisation trend continues to this day.
In the early days of business analytics, the late 50s and the 60s, very few people had any kind of access to operations data beyond accounting and budgets. Only senior management could consult data tools, or what were then known as decision support systems (DSS). These tools were highly constrained by the computational limits of the day, and it's not clear how much of an impact in business DSSs actually had.
Another contemporary term, management information systems (MIS), predates business analytics. The terminology of this era wasn't exactly clear and it certainly hasn't cleared up since — there are countless systems and tools and frameworks that try to accomplish a similar goal, helping people make better decisions.
The point is that at least the desire to see business as a unitary system has a long history. Executives have always wanted to see business as a machine that they can control with data-driven levers. This approach in a sense abstracts away the human element.
Investment in data and data tools has continued to rise ever since. Data is now getting into more hands and is generating business insights at every level of business. The data democratisation trend has resulted in businesses embracing data-driven processes: everybody is benefiting from data analysis, as both author and the audience.
Data stories are a big part of data democratisation. When advanced data science ability is no longer a requirement for understanding, more people can get something out of data. Only basic data literacy is required.
Analytics vendor Qlik offers some great perspective on the history of business intelligence. In their view, business intelligence divides nicely into three generations: centralised, distributed, and democratised.
The first generation, centralised analytics, is based on specialist data analysts running queries and reports to answer specific, pre-determined questions. Results take weeks to assemble, and making any changes is a lengthy process.
The second generation, distributed analytics, represents the improvements that modern BI brought on the scene already some ten years ago. Distributed BI is user driven and powered by interactive tools, such as dashboards and high level query languages. Business analyst have the freedom to explore data at their own pace, extracting answers and results in interactive sessions.
Finally, cutting edge systems represent democratised analytics, where the objective is to go beyond self-service. Augmentation and automation feature heavily in the tools of this generation. Data governance and related controls, and trust in data and models is central. Open access is carefully balanced with privacy and ethical concerns. Tools facilitate data literacy and support non-specialist users. Everybody can get results and build their own data stories — and the tools are often one step ahead of the user.
Gartner and Narrative Science take it one step further, speculating about a Wave 4 of analytics. Narrative Science in particular believe that the future belongs to automatically generated data stories.
Data mountains reach ever higher, and increasingly all of the data is available for analysis. This means that the importance of filtering and distilling information will just keep growing. Communicating insights effectively will be the skill in highest demand.
Data storytelling can be seen as complementary or additive piece to self-service analytics. Data stories will not replace full-on data analysis tools and workflows, those will still be available for the specialists, but for everybody else stories may well become the primary interface.
When individuals are empowered to act on insights, at every level, companies may well see a wave of small, incremental improvements. And small changes compound. At the same time, positive experiences with small changes make greater shifts possible, thinkable.
Persuasive data storytelling is not just about winning buy-in and support for a particular project, but about winning the metagame, so that the business can pursue bigger insights.
The Insights Engine
"Operational skill used to confer long-term advantage. Today those capabilities are tables stakes."
Organising a business around data-driven insights requires a structured approach to the business analytics function.
In a 2016 Harvard Business Review piece, Frank van den Driest, Stan Sthanunathan and Keith Weed write about "insight engines", summing up their experiences doing analytics work at Unilever.
Competitive advantage is founded on a deep understanding of customers and the marketplace — business analytics. Turning data into insights is the name of the game, and the insights engine group plays a central role in this.
By insights engine, the authors refer to "a set of structures, people, and processes that can translate data into actionable strategy." In other words, an insights engine is an augmented analytics group that powers the bus that is supposed to carry the organisation on the road trip across the last mile of business analytics. The insight group, powered by analytics, serves in a high-level advisory role, influencing operations at every level within a company.
The authors give seven operational characteristics of a strong analytics and insights engine group, as well as some further guidance on how to identify the right people to form the team. Effective data storytelling is the main thing.
Data synthesis: Most analytics pipelines begin with data assembly. What really distinguishes solutions in this space is the ease with which data from disparate sources can be linked together, and the scale at which this is possible. It's less about volume, more about connecting the dots. For extra difficulty, throw in some unstructured data.
Independence: For the greatest impact, the analytics function should sit outside other functions like marketing, and report directly to someone in the executive suite. This could be the CEO, or a specialist officer in charge of strategy or experience, or someone focused on data and insights.
Integration: "If insights groups are to help drive strategy, their activities must be aligned during the planning cycle with those of strategic planning, marketing, finance, sales, and other functions." Analytics needs to be part of the conversation, when it comes to where the company is active and how winning is defined.
Collaboration: The engine group should work closely with other functions and customers. Traditional reporting functions are evaluated based on how effective they are as a service provider to other functions, but the insights team needs to be able to form deeper relationships. This means shared goals and an understanding of partnerships within the company.
Experimentation: The analytics group should lead the rest of the company in embracing a data-driven culture of experimentation. This can take many forms.
Future orientation: Traditional business analytics is focused on the past. Today, most people are concerned with having an accurate picture of the present. An insights engine should be driving things forward, and be focused on the future.
Action: The engine group should have a strong "affinity for action". In addition to research and analysis skills, a modern analytics function must be able to communicate, persuade, facilitate and lead.
Whole-brain mindset: The insights team needs people who can think creatively as well as analytically. People who can use both the left and the right side of their brain.
Business mindset: All teams should think about the business impact of their work, and that goes double for the insights engine group. All analytics work should be prioritised and evaluated from a business point of view.
Storyteller mindset: People working with insights should be adept storytellers. The insight engine role is to a large extent about conveying messages through engaging narratives. Impact on strategy is founded on effective, even provocative storytelling.
"It’s such a steep learning curve for business people to grasp data literacy to a level where they can benefit. They need molding and mentoring in a way that they can absorb. If we had really good data storytellers, it would make it so much easier."
Data storytelling is translation. Data stories are interpretations of data encapsulated in a narrative format. This translation effort is worth the trouble, because the new format is such an effective vehicle for insights. Narratives make information easier to consume, and help decision makers absorb insights and initiate appropriate actions.
Bisson et al., writing for McKinsey, believe that the best companies create cross-functional analytics teams. These teams would include committed business representatives, UX people, and data engineers and scientists, but additionally, the teams would have designated analytics translators. These translators would bridge the technical expertise of technical people with the operational expertise of business specialists in other business functions.
Henke et al., also for McKinsey, elaborate on the translator role. In a nutshell, the translator's job is to help ensure that organisations achieve real impact from their analytics initiatives. Translators help ensure that "deep insights generated through sophisticated analytics translate into impact at scale in an organization". Translators help businesses increase the return on investment for their analytics initiatives.
Translators draw on their domain knowledge — within a single line or across the organisation — and help managers identify and prioritise business problems. Translators need a strong sense of what will create value.
Translators take business drivers and present them to the data and analytics specialists. These data professionals then do the analysis or modelling. In the other direction, the translator takes the results, ensures the insights have value, and then communicates the benefits of these insights to business users to drive adoption.
Henke et al. give a set of essential skills and characteristics:
Domain knowledge: Translators must be experts within the industry and within the company. Without this expertise, it's difficult to identify what is valuable. Translators must understand the key operational metrics, the impact on profit and loss, revenue, customer retention, and so on. They also need to understand customers and typical analytics use cases.
Tech fluency: Translators need to be sharp and have affinity with quantitative methods and structured problem solving. Translators need a working understanding of different analytics approaches, but need not grasp all of the details. Interpreting results and identifying potential errors is a key skill.
Project management: Mastery of project management is a must. Translators should be able to direct analytics initiatives from ideation to production and adoption. Translators should have an understanding of the analytics project life cycle as well as knowledge of common pitfalls.
Entrepreneurial spirit: Translators should also have an entrepreneurial mindset. Enthusiasm, commitment and business savvy are all needed when navigating technical, political and organisational roadblocks. This is often culture dependent, and challenging to teach or learn.
Fleming et al., also McKinsey, concur, saying that hiring analytics translators is a great way to "unlock value" from analytics programs. The role is absolutely critical, but often goes overlooked. It's best filled by someone from the business side who can help leaders identify high-impact analytics use cases and work with the analytics team to deliver actionable solutions. Translators play a vital role in generating buy-in and in scaling solutions from pilots across the organisations.
The authors all believe that training existing employees is the best option for finding translator talent. External hires would take years to develop the necessary deep company knowledge. Unfortunately there aren't many options for certification in this area yet, but that will likely change in the future.
The consultants recommend setting up internal translator academies to train candidates for this roles, as this unique combination of skills is hard to find.
Convincing others about the merits of a data-driven insight is central to data storytelling. In particular, winning support from executives and senior managers is essential, if company scale changes are being proposed. The bigger the insight, the more critical it is to communicate it well all the way to the top.
McShea, Oakley, and Mazzei write in a 2016 HBR piece about the role that CEOs can play in supporting analytics programmes. "Analytics forces changes on the C-suite that the CEO has to anticipate and manage, but many don’t." Lacklustre return on analytics investment is often just the tip of the iceberg.
The authors give several indicators that something is wrong with the analytics programme. If senior managers are frustrated with analytics, the analytics pipeline has long delays, or the analytics function requires frequent restructuring, then something is definitely going on. Churn in the C-level post responsible for analytics and clear over-reliance on consultant help is similarly problematic.
The CEO should appoint the right analytics leader, someone with "a vision for how analytics could drive the company to a brighter, perhaps radically different, future". A big part of the job is to create "an environment of discovery", such that data can shape the future of the company. Rapid innovation and experimentation must be rewarded and supported.
The main thing for the CEO is to actively manage C-suite dynamics. Fully integrating data and analytics into a business may require that executives rethink some of their hard won mental models, which may be a difficult transition. Executives may be attached to the status quo, and work against to undermine analytics initiatives — in fact they stand to gain, if they manage to discredit the "innovators".
"While CEOs typically understand the transformative potential of big data/analytics, they often do not consider the flip side of the coin — that analytics efforts unleash forces within an organization that can threaten the analytics program itself."
McShea et al. repeat their warning in another HBR piece: "Efforts to adopt analytics upset the balance of power in the C-suite, and this shift often had a negative impact on analytics initiatives." Driving major innovations with analytics often proves harder than many executives expect. "Further study of the less-successful cohort reveals that leadership issues were often at the heart of the problems."
Weak leadership manifests as a range of issues. Commitment to analytics disrupts the C-suite equilibrium, and without a natural owner, multiple executives can compete hard to own this new capability. Those who "lose out" may feel vulnerable and protective, which leads to inefficiencies and missed opportunities in the functions not controlled by the analytics "winner".
"In all too many cases, the CEO devoted little time to trying to manage this dynamic," the authors write. The CEO must play a leading role in establishing the analytics function in the organization. CEOs should anticipate the reaction and speak openly about the impact that upcoming changes will have. Transparency "moves the dialogue above the table."
CEOs should identify the executives most vested in the status quo and proactively manage this resistance. Managing the power shifts brought on by analytics is a major challenge for the CEO, but it must be done right, or the analytics mission will never succeed in delivering competitive advantage.
In short, the CEO must step up. The necessary upgrades and transformative change require clear gestures and intent from senior leadership. "Discovering and creating the future requires many eyes and total collaboration."
A 2016 McKinsey survey on the need to lead similarly emphasises the integral role that executives have in making analytics initiatives a success. "The results suggest that the biggest hurdles to an effective analytics program are a lack of leadership support and communication, ill-fitting organizational structures, and troubles finding (and retaining) the right people for the job."
Simply put, high performing companies have CEOs driving the analytics programmes. In lower performing companies this task has been given to a champion a step below the C-suite. The difference in impact is proportional.
"For so many organizations today, technology is the business. Technology needs to be understood as a critical enabler in every part of the organization from the front line to the back office. It creates new value by crunching data to deliver new insights, it spurs innovation, and it disrupts traditional business models."
Speaking Truth to Power
Communication needs to be equally strong on the other side of the equation. Ashford and Detert write, for HBR, that managers should learn how to sell their ideas up the chain of command. Organisations prosper only if the middle ranks feel empowered to identify and promote ideas when there is need for change.
"Studies show that senior executives dismiss good ideas from below far too often," the authors write. "If [executives] don’t already perceive an idea’s relevance to organizational performance, they don’t deem it important enough to merit their attention. Middle managers have to work to alter that perception."
To do this persuasion effectively, middle managers need to have a good relationship with their audience and feel psychologically safe to bring things up. Middle managers need to believe that someone will hear them out, if they bring forward an issue they believe in.
Ashford and Detert give a selection of tactics to explore when selling ideas up. Naturally, data stories can be an effective component in this.
Brian Eastwood, reporting for MIT Sloan, writes about the importance of effective communication, especially when it comes to the results of a data initiative. "One of the biggest challenges in data analytics is presenting results in a way that’s meaningful to people who aren’t data scientists."
It all comes down to trust. The stakeholders have to be able to trust your data. They trust you to only bring forward results that you have identified to be meaningful. Any models need to be easy to interpret, especially when it comes to people without a background in data analytics. Data storytelling can be an effective way to win audience interest in how a model solves a problem.
When proposing changes, one of the main challenges is addressing and overcoming any resistance to the proposal. It's a good idea to anticipate stakeholders’ cognitive biases, such as knee-jerk reactions or the silo effect, and try to overcome them by asking the right questions and establishing a framework for making decisions, Eastwood writes. "Analytics is more people-related." How data results get presented is key.
Hal Gregersen writes, for HBR, about another challenge in communicating upwards: radio silence. "Persistent CEOs almost always get the information they request. It might not arrive as fast as they’d like, but eventually it gets there. Their bigger problem is getting information they haven’t demanded because they don’t know to ask for it."
"If you’re a leader, you can put yourself in a good-news cocoon," Gregersen writes. He recommends focusing on insightful questions, questions that can light up "the territory of unknown unknowns". For CEOs, Gregersen recommends the question: "If you were in my job, what would you be focusing on?"
Another tactic is to insist on brutally honest reports, delivered raw and unadulterated. The idea is to provide the C-suite with a 24/7 early-warning information systems that alerts the right people when something isn't working.
"The ingredient for the effective use of data and analytics that is in shortest supply is managers’ understanding of what is possible."
Data storytelling is the craft of presenting data-driven insights in a compelling narrative form in order to drive action and change within an organisation.
Data storytelling seeks to reinvent our relationship with data by focusing on the way we communicate with data.
Facts and figures alone won’t influence decisions or move people to act, but well-crafted data stories can and will.
Stories are important, because they get through to people on the emotional side.
Without context and empathy in the presentation, insights will not get leveraged in decision making.
Data storytelling goes a step beyond dashboards. The narrative form helps people absorb information beyond what visualisation can do alone.
Many business intelligence vendors are looking at data storytelling, and many believe that augmented analytics, AI powered analytics automation, is the answer.
Data storytelling is fundamentally an exercise is moving an insight from one brain to another. To this end, pretty much anything goes; there are no fixed rules.
The main thing is to keep things simple.
One-sided expositions lead to ineffective conversations, as people don't have time to process the news before the wrap-up discussion.
Always have a "speedthrough" version ready.
A good data story has three parts: setup, conflict and resolution.
Be prepared, but not too prepared.
Self-service analytics won't be enough in the future. Those tools will not go away entirely, but business analytics will be disrupted by augmented intelligence and data stories.
When advanced data science ability is no longer a requirement for understanding, more people can get something out of data.
Organising a business around data-driven insights requires a structured approach to the business analytics function.
Data storytelling is translation.
Winning support from executives and senior managers is essential, if company scale changes are being proposed. The main thing for the CEO is to actively manage C-suite dynamics.
It all comes down to trust: the audience has to trust the data, and the storyteller has to believe that the audience is interested in the story.
Andreessen, Marc. 2011. Andreessen Horowitz. Why Software Is Eating the World.
Ashford, Susan J. and James R. Detert. 2015. Harvard Business Review. Get the Boss to Buy In.
Avidon, Eric. 2020. TechTarget. Data-driven storytelling opens analytics to all.
Avidon, Eric. 2021. TechTarget. Gartner predicts data storytelling will dominate BI by 2025.
Berinato, Scott. 2019. Harvard Business Review: Quick Study. Telling Stories with Data in 3 Steps (Quick Study).
Bersin, Josh. 2016. Harvard Business Review. Good Presentations Need to Make People Uncomfortable.
Bisson, Peter, et al. 2018. McKinsey & Company. Breaking away: The secrets to scaling analytics.
Brown, Brad and Josh Gottlieb. 2016. McKinsey & Company. The need to lead in data and analytics.
Browne, Nolan. 2014. Harvard Business Review. Presentation Tools That Go Beyond “Next Slide Please”.
Campbell, Josh. 2018. TEDxMemphis. Seven Keys to Good Storytelling.
Coleman, John. 2015. Harvard Business Review. Create a Conversation, Not a Presentation.
Davenport, Thomas H. 2015. Harvard Business Review. 5 Essential Principles for Understanding Analytics.
Dykes, Brent. 2020. Forbes. Why Data Storytellers Will Define The Next Decade Of Data.
Eastwood, Brian. 2017. MIT Sloan. Talking to your boss about data.
Fairfield, Hannah. 2012. The New York Times. Driving Safety, in Fits and Starts.
Fleming, Oliver, et al. 2018. McKinsey & Company. Ten red flags signaling your analytics program will fail.
Gregersen, Hal. 2017. Harvard Business Review. Bursting the CEO Bubble.
Hagerty, John. 2019. Every Picture Tells a Story, Don’t It? But Is It the Right Story?.
Haroz, Steve, et al. 2015. Kellogg Insight, Northwestern University. Data Visualization: More Than Pretty Pictures.
Henke, Nicolaus, et al. 2018. McKinsey & Company. Analytics translator: The new must-have role.
Kendall, Stu. 2020. Narrative Science. Welcome to the Third Wave of Analytics.
Little, Cinny. 2018. Forbes. Drive More Impact From Data And Analytics With Insights Storytelling.
Little, Cinny, et al. 2020. Forrester. Use The Five-Step Playbook For Data And Insights Storytelling To Drive Actions That Matter.
McShea, Chris, et al. 2016. Harvard Business Review. How CEOs Can Keep Their Analytics Programs from Being a Waste of Time.
McShea, Chris, et al. 2016. Harvard Business Review. The Reason So Many Analytics Efforts Fall Short.
Ostrowski, John. 2019. LADDER. Data Storytelling: Best Practices and Visualization Tips To Create Remarkable Reports.
Poindexter, Will. 2019. Bain & Company. Bridging the Leadership Gap Between Tech and Business.
Qlik. 2021. Business Intelligence.
Shield, Cassidy. 2020. Narrative Science. Data Storytelling and Augmented Analytics Stand Out at Gartner’s D&A Summit 2019.
Shield, Cassidy and Brent Dykes. 2020. He Wrote the Book on Data Storytelling | Data Stories Deep Dive Episode 1.
Somers, Meredith. 2020. MIT Sloan. Our top 5 “Working Definitions” of 2020.
Stackpole, Beth. 2020. MIT Sloan. The next chapter in analytics: data storytelling.
Van den Driest, Frank, et al. 2016. Harvard Business Review. Building an Insights Engine.
Victor, Bret. 2014. worrydream. A note about “The Humane Representation of Thought”.