Skip to content

Data-driven storytelling opens analytics to all

    Data storytelling, because it interprets and explains data, extends business intelligence to business users and not just those trained in data analysis.

    Data-driven storytelling has the potential to revolutionize analytics.

    One of the great challenges of analytics has been making it accessible to more than just the trained experts within an organization, the data analysts who understand how to interpret data and use it to make informed decisions.

    And just as visualizations helped make data more digestible a decade or so ago and augmented intelligence is making analytics platforms easier for untrained users to navigate, data-driven storytelling can put business intelligence in the hands of a wider audience.

    But unlike data visualizations and AI, technologies that only marginally extend the reach of analytics, data-driven storytelling can have a wider impact in enterprises.

    Data storytelling, simply, is an automatically generated explanation of data. It’s the story of the data under analysis, that interpretation that if left to someone without an expertise in data analysis can be dangerous. It’s that story put in a narrative form rather than just straight analysis of the data.

    “Data storytelling is what you say when you’re actually trying to understand what’s happening in the data and make a decision off of it,” said Nate Nichols, chief scientist at Narrative Science, a data storytelling software vendor.

    For example, Nichols continued, if someone comes home and sees a spilled glass of water on the kitchen counter and the wet footprint of a cat leading away from the water, they have a data set.

    “That’s what you get from a spreadsheet or a dashboard,” Nichols said. “But you don’t make a decision based off of that. You develop an interpretation of what happened, you tell a story — the cat came in, tried to drink, knocked over the water and ran out. It’s the story that helps you make the decision about how to keep the cat away in the future.”

    In a business sense, data-driven storytelling, for example, can be the explanation of sales figures in a report or dashboard.

    Rather than just present the numbers and leave the interpretation up to the user, data storytelling platforms break it down and put into a written narrative that total sales in a given week were $15 million, which was up 10% over the week before and up 20% over the weekly average. Meanwhile, the sales figures include 100 deals with a certain employee leading the way with eight, and the overall increase can be attributed to seasonal factors.

    A Narrative Science data story about the sales figures highlights the most relevant numbers in bold, creates a simple bar graph, and situates a block graphic below a bold headline over the narrative. A traditional spreadsheet would leave it to the user to interpret the same information presented in rows of numbers.


    A sample data story from Narrative Science describes an organization’s sales bookings.

    And while data-driven storytelling has the potential to open up analytical analysis to the masses, it isn’t merely for the benefit of those untrained in the language of data. Even those with backgrounds in data science can struggle to find the meaning within data that can lead to action.

    “As a trained analyst myself, data was always a means to an end,” said Lyndsee Manna, senior vice president of business development at Arria NLG, a natural language generation vendor. “But I, as a human, had to wrestle to extract something that was meaningful and could communicate to another human. The shift to data storytelling is that I don’t have to wrestle with the data anymore. The data is going to tell me. It’s knowledge automation.”

    The Psychology

    Human beings understand stories.

    From the earliest cave dwellers telling stories with pictures through the present day, people have used stories to convey information and give it context. Analytics, however, has largely lacked that storytelling aspect, missed out on the power a story can have. Even data visualizations don’t tell stories. They present data in an easily understandable format charts, graphs and in artful ways, but they usually don’t give the data meaning in a richer context.

    And that leaves countless business users out of the analytics process. Data storytelling changes that. “It gives information context, it gives it purpose, and makes it more memorable and understandable,” said Donald Farmer, principal at TreeHive Strategy. “For that reason it’s very fundamental psychologically. Storytelling is essential. In a sense, data storytelling is nothing new because whenever we exchange data we do it with implicit stories. But data storytelling as a practice is emerging.”

    Similarly, Sharon Daniels, CEO of Arria NLG, said data-driven storytelling could revolutionize analytics because of the way humans react to narratives. “If you follow how we evolved as human beings we started in caves with drawings and communicating with visuals, and then language came about and opened up our world, and the technology world is mirroring that,” she said. “It’s a very interesting thing to see the storytelling parallels. The language component and the storytelling is universal.”

    Meanwhile, because of the way people use storytelling to give meaning to information and because of the technology now being developed by data storytelling vendors that automatically generates a story to accompany data anyone in an organization can use data to inform decision-making.

    According to Dave Menninger, research director of data and analytics research at Ventana Research, only between 20% and 40% of employees within most organizations use analytics in their jobs.

    “Data-driven storytelling can expand the reach of analytics. The promise of storytelling is that we’ve been stuck at this level — pick a number — of penetration of BI into an organization, and we have the opportunity to achieve close to 100% with data storytelling,” Menninger said.

    In an informal sense, data-driven storytelling already permeates entire organizations. When a CEO speaks about earnings, for example, they start with hard numbers and then contextualize those numbers with a story. New technology, however, can extend the reach of analytics in a more structured way.

    “What’s happening now is these specific technologies that are being developed to support data storytelling are coming out, and they will [eventually] reach 100% of the organization,” Farmer said. “That’s why it’s exciting in a technology sense. We’ve finally got a technology that actually can genuinely reach everyone in some way.”

    The Technology

    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.

    Vendors such as Alteryx and Teradata specialize in data management, loading the data and structuring it. Others such as Tableau and Qlik are specialists in the business intelligence layer, the presentation of the data for analysis. And still others, including software giants IBM and Oracle, enable each aspect of the analytics process.

    Now, a crop of vendors has emerged that specialize in data-driven storytelling, taking that data that’s gone through the entire pipeline and giving it meaning.

    Narrative Science, though founded only 10 years ago, is one of the veterans. Arria NLG, which offers a suite of natural language generation tools in addition to its data storytelling capabilities, is another that’s been around for a while, having been founded in 2009. And now startups like Paris-based Toucan Toco are emerging as data storytelling gains momentum.

    Meanwhile, longstanding BI vendors are also starting to offer data storytelling tools. Tableau introduced Explain Data in 2019, and Yellowfin developed Yellowfin Stories in 2018.

    “Everyone wants everyone to be able to make data-driven decisions and not have to have an analytical background or have their own analysts,” Nichols said. “But for people that aren’t analysts and that are just trying to understand the story and use that to guide their decision-making, that last mile is the hurdle.”

    According to Nichols, Narrative Science’s data stories are generally short and to the point, often only a paragraph or two, though they have the potential to be longer. Arria NLG’s stories can similarly be of varying length, depending on the wants of the user.

    “Whether you know about data and excel at BI or whether you don’t, data can feel very overwhelming,” Manna said. “The biggest thing [data storytelling] gives to humanity is to lift that feeling of being overwhelmed and give them something — in language — they can comprehend quickly. The gift is understanding something that either would have taken a very long time or never to understand.”

    Data stories are generally the final phase of the analytics process rather than embedded throughout. When BI vendors offer their own data-driven storytelling tools, they generally provide the opportunity to embed stories at points along the data pipeline, but that can be tricky, according to Farmer. If the tools are introduced too early in the process, they can influence the outcome rather than interpret the outcome.

    “You have to be very careful with data storytelling,” Farmer said. “For me, data storytelling has to be focused on a single subject.”

    In addition, he said, it’s important to understand that data storytelling doesn’t completely replace analysis. The stories produced by data storytelling platforms are linear, and the real world is far more meandering.

    The Future

    Unlike most new technologies that start off in rudimentary forms and develop over long periods of time, data storytelling platforms already deliver on the promise of providing narratives that contextualize data and help the decision-making process. However, they have the potential to do more.

    Data-driven storytelling platforms don’t yet know their users. They can analyze data and craft a narrative based on it, but they don’t yet have the machine learning capabilities that will lead to personalized narratives.

    “It really should be personalized explanation of the analysis and personalized instructions on what to do based on the observations,” Menninger said. “Many vendors are at the point of explaining, and those explanations may be somewhat personalized for the region or product you’re responsible for, but few vendors have gotten to the point where they’re offering instructions.”

    He added that with machine learning, the tools eventually will recognize that a person might look at a certain monthly report or dashboard and then follow up by doing the same thing each time. But people with similar roles within the organization might do something different after they look at that same report or dashboard, so the software will recommend that perhaps the first person ought to be doing something different after looking at the data.

    Daniels, likewise, said personalization is an important part of the future of data-driven storytelling. “I would say the ultimate data storytelling would be hyperpersonalized, predictive analytics that is telling me not only what happened and trends but is also telling me what to look out for and what could happen in the future,” she said. “It’s bringing more predictive analytics into the data storytelling, and we’re not far off from that.”

    Beyond personalization, data storytelling platforms will likely evolve to be more proactive, according to Nichols. Now, the platforms require users to open their reports and dashboards and request the narrative. “Part of data storytelling is understanding when a story needs to be told,” Nichols said. “And when I think of a perfect vision for data storytelling, it’s the system being proactive. It’s telling you when there’s a story you need to hear.”

    And the same is true conversely, he added, when there’s nothing new of note and there’s no reason to generate a new story. No matter what the future holds, however, data-driven storytelling tools will always be about extending the reach of analytics to a broader audience, and for the first time, potentially everyone.

    Use Cases of Big Data Analytics in Real World

    The most valuable item for any company in modern times is data! Companies can work much more efficiently by analyzing large amounts of data and making business decisions on that basis. This means that Big Data Analytics is the current path to profit! So is it any surprise that more and more companies are gradually turning towards a data-based business model?

    Big Data Analytics is much more objective than the older methods and companies can make the correct business decisions using data insights. There was a time when companies could only interact with their customers on one in stores. And there was no way to know what individual customers wanted on a large scale. But that has all changed with the coming of Big Data Analytics. Now companies can directly engage with each customer online personally and know what they want.

    So let’s see the different ways companies can use Big Data Analytics in the real world to improve their performance and become even more successful (and rich!) with time.

    1. Companies use Big Data Analytics to Increase Customer Retention

    No company can exist without customers! And so attracting customers and even more importantly, retaining those customers is necessary for a company. And Big Data Analytics can certainly help with that! Big Data Analytics allows a company to observe customer trends and then market their products specifically keeping their customers in mind. And the more data that a company has about its customer base, the more accurately they can observe customer trends and patterns which will ensure that the company can deliver exactly what its customers want. And this is the best way to increase customer retention. After all, happy customers mean loyal customers!

    An example of a company that uses Big Data Analytics to Increase Customer Retention is Amazon. Amazon collects all the data about its customers such as their names, addresses, search history, payments, etc. so that it can provide a truly personalized experience. This means that Amazon knows who you are as soon as you log in! It also provides you product recommendations based on your history so you are more likely to buy things. And if you buy lots of things on Amazon, you are less likely to leave Amazon!

    2. Companies use Big Data Analytics to create Marketing Campaigns

    How can a company reach new customers? Marketing campaigns! However, if a great marketing campaign can get customers for a company, a poor marketing campaign can make a company lose even its existing customers. And so Big Data Analytics is necessary to analyze the customer base and understand what people want so that the marketing campaign is successful in converting more people. This can be done by monitoring the current online trends, understanding customer behavior in the market and then cashing on that to create a successful marketing campaign.

    An example of a company that uses Big Data Analytics to create Marketing Campaigns is Netflix. Have you noticed that as soon as you open Netflix, they have movies and series marketed specifically for you? They do this by collecting data on your watching habits and search history and then providing targeted adverts. So if you have been watching mystery movies recently, that’s what you will be recommended in the future as well!

    3. Companies use Big Data Analytics for Risk Management

    A company cannot sustain itself if they don’t have a successful risk management plan. After all, how is a big company supposed to function if they cannot even find risks ahead of time and then work to minimize them as much as possible? And this is where Big Data Analytics comes in! It can be used to collect and analyze the vast internal data available in the company archives that can help in developing both short term and long term risk management models. Using these, the company can identify future risks and make much more strategic business decisions. That means much more money in the future!!!

    An example of a company that uses Big Data Analytics for Risk Management is Starbucks. Did you know that Starbucks can have multiple stores on a single street and all of them are successful? This is because Starbucks does great risk analysis as well as providing great coffee! They collect data like location data, demographic data, customer preferences, traffic levels, etc. of any location they plan to open a shop and only do it if the chances of success are high and the associated risk is minimal. So they can even choose locations that are close together as long as there is more profit and less risk.

    4. Companies use Big Data Analytics for Supply Chain Handling

    The supply chain begins with the creation of raw materials and ends at the finished products in the hands of the customers. And for large companies, it is very difficult to handle this supply chain. After all, it can contain thousands of people and products that are moving from the point of manufacture to the point of consumption! So companies can use Big Data Analytics to analyze their raw materials, products in their warehouse inventories and their retailer details to understand their production and shipment needs. This will make Supply Chain Handling much easier which will lead to fewer errors and consequently fewer losses for the company.

    An example of a company that uses Big Data Analytics for Supply Chain Handling is PepsiCo. While the most popular thing sold by PepsiCo is Pepsi of course, did you know they sell many other things like Mountain Dew, Lays, 7Up, Doritos, etc. all over the world! And it is very difficult to manage the Supply Chain Handling of so many things without using Big Data Analytics. So PepsiCo uses data to calculate the amount and type of products that retailers want without any wastage occurring.

    5. Companies use Big Data Analytics for Product Creation

    All companies are trying to create products that their customers want. Well, what if companies were able to first understand what their customers want and then create products? They would surely be successful! That’s what Big Data Analytics aims to do for Product Creation. Companies can use data like previous product response, customer feedback forms, competitor product successes, etc. to understand what types of products customers want and then work on that. In this way, companies can create new products as well as improve their previous products according to market demand and become much more successful and popular.

    An example of a company that uses Big Data Analytics for Product Creation is Burberry, a British luxury fashion house. They provide luxury with technology! This is done by targeting customers on an individual level to find out the products that they want and focusing on those. Burberry store employees can also see your online purchase history and preferences and recommend matching accessories with your clothes. And this makes a truly personalized product experience which is only possible with Big Data Analytics.