The Real Story Behind Big Data

It was a dark and stormy night

Thanks to the promise of business analytics, seeing data as the new oil is all the rage these days, which is why something like 85 per cent of Fortune 500 companies are investing in big data initiatives. As The Economist recently put it, “Data are to this century what oil was to the last one: a driver of growth and change.” But as The Economist also noted, successfully mining data really isn’t like extracting, refining, valuing, and trading any previous resource because it “changes the rules for markets and it demands new approaches from regulators.” It also requires deeper thought from users, which is why, despite the exponential growth and interest in business analytics, there are big question marks over the extent to which organizations actually realize value from analytics.

It is easy to dismiss the potential of analytics as yet another version of technological snake oil. As TechCrunch recently noted, “Clicking through pages of ‘unlock the value of your big data!’ advertorials, a cynic might suspect that the best (and perhaps only) method of deriving value from big data is to go into the business of telling people how to get value from their big data.” But the real problem is that most organizations simply do not know how to take advantage of big data. “All that’s happened is that technological innovations in data handling capability (made by companies like Google to deal with the scale and complexity of Web 2.0) temporarily leapt ahead of our progress in learning how to apply them—progress we make through experimentation.”

We agree that there is a user problem getting in the way of success. Through our research and consulting work, we have come across many examples of expensive analytics projects failing to deliver, even when state-of-the-art tools and highly competent data scientists are in place. In most cases, the data analysis is sound and the predictive models are great, but the resulting insights are not acted on because they require change in the organization. So nothing gets done

Why so? In order to really leverage the power of analytics, the capability has to be “woven” into the stories of the organization and its people. In our hyper-connected world where data is king, storytelling may seem old-fashioned. But we remember stories, not spreadsheets.

“Drawing on Romeo and Juliet, the story began with images of a harmonious relationship between departments, until the implementation of a new HR management system went horribly wrong.”

Storytelling is truly universal through all of recorded history, from oral storytellers and early “rock art” in hunter-gatherer tribes to the proliferation of writers churning out books, TV shows, and movies today. It is universal, transcending all cultures across all continents, with some of the most ancient cultures writing text in Chinese, Egyptian, Greek, Latin, Sanskrit, and Sumerian.

Senior managers may feel that crafting a story around the data is a pointless and laborious effort—that the facts alone are enough to initiate the desired change. Unfortunately, this opinion is based on the flawed notion that business decisions are solely based on logic and reason. A multitude of experiments from the field of behavioural economics clearly prove that emotion, not rational thinking, is what determines the decisions we make. And very few things stir our emotions quite like a good story. Data may hold tremendous amounts of potential value, but if an insight isn’t understood and isn’t compelling, no one will act on it and no change will occur.

Throughout time, storytelling has proven to be a powerful delivery mechanism for sharing insights and ideas in a way that is memorable, persuasive, and engaging. A good example of modern-day storytelling is the popular TED conference series and its slogan of “Ideas Worth Spreading.” As pointed out in a Forbes column by Carmine Gallo—author of The Storyteller’s Secret: From TED Speakers to Business Legends, Why Some Ideas Catch on and Others Don’t—Facebook COO Sheryl Sandberg’s now famous 2010 TED talk, “Why We Have Too Few Women Leaders,” launched a movement by starting with:

I left San Francisco, where I live, on Monday, and I was getting on the plane for this conference. And my daughter, who’s three, when I dropped her off at preschool, did that whole hugging-the-leg, crying, “Mommy, don’t get on the plane” thing. This is hard. I feel guilty sometimes. I know no women, whether they’re at home or whether they’re in the workforce, who don’t feel that sometimes.

According to Gallo, if you conduct an analysis of the most popular 500 TED Talk presentations, you’ll find stories make up at least 65 per cent of their content. Combining data with a good story, of course, is a delicate balancing act. Too much story and not enough data and you’ll quickly be dismissed as lacking legitimacy. Too heavy on the numbers and nobody tunes in, as we’ve seen in our experiences.

A medical technology company we worked with conducted an extensive organizational network analysis, mining e-mail patterns and employee surveys, to determine the level of collaboration between its various business units. The resulting visual and mathematical data revealed a massive rift between the IT and HR personnel. Management concluded that missed efficiencies and suboptimal HR systems were the consequence of this lack of collaboration. The operations manager who led the network analytics project realized that just presenting the visualizations and numbers to members of IT and HR was unlikely to result in any constructive change. Instead, he used the network analytics to enhance the story he was going to tell. On reporting back the findings to IT and HR, he followed the Freytag’s pyramid approach—the dramatic arc of exposition, rising action, climax, falling action, and dénouement used in most movies and plays—to structure his story.

Drawing on connotations from Romeo and Juliet, the story began with images of a harmonious relationship between the departments, until the implementation of a new HR management system went horribly wrong. The heads of the IT and HR families blamed each other (which is what actually happened two years earlier in the company) and relations between the two soured. The network analytics were then produced as evidence of the clear rift. To reignite collaborations, the operations manager turned to the “falling action” element of the story, where the conflict begins to thaw in the future. A small few heroes from both groups begin to seek simple advice from each other, which grows into collaborating together on a proposal for a more streamlined online recruitment system, from which introductions and other collaborations are spawned between the two departments and more successful projects emerge. Pulling in the network analytics again, the operations manager produced visuals illustrating how the rift would heal over time as a result of the actions of the heroes, with the intimation to the audience to become one of the heroes. When analytics were woven into the story, the change required to advance the organization became much more tangible.

While it may seem intuitive to weave analytics into and around the stories of your organization, our research and consulting experience shows that this can be quite challenging. Ineffective and sometimes damaging use of stories and analytics will emerge, unless a number of steps are carried out:

  • Hire and train good analytics storytellers: Data scientists are in huge demand in the labour market, but very few companies are thinking about storytelling skills. By all means, ensure your company gets the data science skills needed, but don’t miss the opportunity to complement those deep analytics skills with team members who are more gifted at narrative—who are close to the business context and great at helping others see the big picture that the data underlines. While the crafting of a good story may come naturally to some, it is also a skill that can be learned, even by those with the deepest of analytics specializations. But it takes work and investment. Leaders also need to recognize that there may be people in their team who are better at crafting a story with analytics, and they should be chosen to deliver the story to an audience.
  • Complement storytelling and analytics: Storytelling techniques can be complemented with cold data to ensure that the richness of the story is corroborated with the credibility and objectivity of supporting data. At Fidelity Investments, the ITEC team was charged with leading the firm toward a standardized approach to planning, building, and shipping software across its federated business units. By combining visualization of both the current state and progress across each business unit with a narrative of benefits to employees and the firm, the ITEC team was able to accelerate progress, mobilize the organization towards the goal, and get thousands of technologists to embrace and adopt the standards. An additional benefit of the visualization techniques was how they enabled comparison of progress to drive a dialogue across technology leadership, which steered the story in the right way. Those who were able and willing to trailblaze could reassure other leaders on the benefits from standardization. Analytics technology such as eye-tracking and emotion-sensing technology can also help build rich stories. For example, in proposing to increase the speed limit at which U.K. motorists can travel through highway roadworks, Highways England did not just rely on personal stories of drivers stuck in traffic congestion, but instead conducted a study using heart-rate monitors, GPS trackers, and dash cameras to measures reactions. Drivers are less stressed when driving at 60 miles per hour than 50 miles per hour. The combination of the analytics and the story tells a much richer story than either one alone.
  • Tailor analytics to fit the story: People get excited about analytics, and strive to use the latest and greatest tricks and features. Instead, one needs to think about the “stories” of the organization and people in it. If, for example, an executive arrives from a run at 8:55 each Monday morning, and has five minutes while grabbing coffee before going into the weekly resource-allocation meeting, then the analytics report must fit that scenario. A 20-page analytics report, even if of high quality, will not be read, whereas a one-page, three-point analytics summary, even if crude and rudimentary compared to the other, is more likely to be used.
  • Ensure appropriate authorship: Stories are an embodiment and physical representation of their author, and so carry their positive and negative traits. Choose authors carefully, as the analytics capability will only be as powerful as those creating the story. Some are highly creative, visionary, “next-gen,” and outside-the-box thinkers, while others are very limited creatively, live in the past, love the status quo, and don’t want anything that would require a change of mindset and a bit of extra work. Many authoring strategies can be employed. One trustworthy person may author all stories, authorship can be delegated to those with the niche expertise specific to each story, or indeed the customers or users can write the stories. Regardless of strategy, an effective peer-review process to evaluate analytics stories is critically important.
  • Ensure analytics stories capture your dark side: Analytics capabilities are often built on idealism but fail to acknowledge an organization’s informal, messy realities. There are patterns of power and domination at play, and staff “weaponize” data—using and manipulating analytics to support their own agenda rather than that of the organization. While getting people to relay their darker side—their dislikes, insecurities, and fears—is inherently difficult, it allows analytics designers to create something that addresses and embraces those traits, and so ultimately is more likely to be used.
  • Make analytics stories live: Ensure that the weaving of analytics and stories is continuous in your organization, rather than a one-off or occasional initiative. For many reasons, stories degrade very quickly over time, which rapidly and radically diminishes their power as a communication tool. Firstly, the story has a “glory period,” where it is often richly described, discussed, and debated when it is first created. However, novelty quickly wears off, and the level of debate and discussion soon drops. Secondly, the person who created the story and those that were there at the time it was told often move on. The rich, tacit, non-documented nature of stories that are not on paper can be lost on staff that were not there when they were relayed. Finally, the story describes a context, a persona, or a need that changes over time. Therefore, any story will suffer from natural degradation, as it captures less and less of the continuously evolving and emerging stories of the organization.

Storytelling is not a panacea and not all stories are good. A bad story is confusing and boring, and can cripple the valuable insights painstakingly gathered from an analytics initiative. But with effort, a story enhanced with analytics can inspire real organizational change.

About the Author

Kieran Conboy is a Professor in Information Systems and leads the Lero research group at NUI Galway. He previously worked for Accenture Consulting and the University of New South Wales in Australia.….Read Kieran Conboy's full bio

About the Author

Eoin Whelan is a Lecturer in Business Information Systems at the National University of Ireland, Galway. He is also a visiting professor at the Institute d’Economie Scientifique et de Gestion….
Read Eoin Whelan's full bio

About the Author

Seán Morris is Senior Vice President at Fidelity Investments and leads the IT Enablement and Operating Model group. Seán has been involved in technology leadership, strategy, transformation….
Read Seán Morris's full bio

Leave a Reply

Please submit respectful comments only, including full name, professional title, and contact information (only name and title will be posted). Required fields are marked *