What Comes After Big Data?

Predictive modeling is but an early step; we must see beyond the fleeting ability to increase underwriting profit or fast-track a claim process.

The force of transformation in our technological age is undeniable, unpredictable, rapid and without controls to slow or stop. No industry can freeze a convenient moment in time when its commodity has high value that is safe from competitive disruption and in perfect alignment with technology. Any and every business can be blindsided by a competitor’s next-generation upgrade of IT, or an upstart’s reinvented consumer acquisition and interaction experiences. What is new today becomes old in a flash. In the risk and insurance industry, investment is booming in predictive analytics and big data. Many proponents envision the death rattle of stodgy experience mod rating, which would give way to “Moneyball” fantasies flush with evergreen underwriting profits. While Moneyball fantasies may pan out for now, our industry cannot control the genie emerging from this bottle. I suggest we consider when and how big data might mature as a cheap ubiquitous commodity and how to hone the next logical step that capitalizes on its inevitable demise. First, we must accept that the devaluation of predictive analytics is imminent, whenever that comes. Consider these questions: --Will analytics still create any underwriting advantage when all companies are applying similar models? --How will the “smart money” know when to stop huge investments in model-building? When 900 data points show no more appreciable value than 400? When the burdensome collection of data at the adjuster’s interface limits, dumbs down, dehumanizes and fast-tracks the front-line adjusting operation so as to, ironically, become a detriment to claim outcomes in and of itself? See Also: Competing in an Age of Data Symmetry --What happens when the first major broker or marketing interest cracks the dam and applies analytics as a loss-leader to fish for clients or tangentially grow a related market share? For example, offering to analyze a prospect’s work comp for free as part of winning a lucrative global property program. Can you beat the rush as more consumers expect predictive analysis “freebies” as part of the entry expense for winning customer contracts? --How soon will some website’s appetite for “click-bait” mean that it offers free, robust, on-line predictive analytic calculators simply to build email lists of potential WC customers? --What if government interests unleash the ability to apply top-notch WC analytics on an open-source employer platform for the good of the state? Can self-use, cost-saving analytics become a public “right” and not a paid-for “privilege”? Today’s reality is simple: Information is vast, easily accessible and free. This fact not only foretells the demise of the value of big data in our industry, but it also instigates the next step in creating opportunity. This next step will arise from the changing nature of higher education and future job seekers. I was recently privileged to hear a talk by the headmaster of an esteemed college preparatory school, who espoused a necessary wholesale change in education. His premise: There is no longer any value in teaching students facts and information because all of it is available and accessible for free. He considers it educational malpractice to make students learn facts. He has shifted a good part of his school curriculum to project-based learning. Student teams are presented with issues or situations and create solutions or new perspectives that open higher possibilities. One of the project teams tackled the challenge of cross-teaching Mandarin and English languages. Their research discovered that the Chinese have a passion for U.S. basketball. The team produced a video of instructional interactive basketball drills that taught language during the real-time experience of following drill instructions. Their first module is now actually being used in China to support prospective students interested in American schools. The headmaster jokingly said his school may have to forego non-profit status to look for investor money and make the concept a complete language package. Mind you, these creators are teenagers with no real budget who were able to use the Internet and common technology to research, design and produce this valuable product and change notions of language-learning. The bottom line is that future employee talent will not care to know facts but will find its highest value in being able to ignore the conventional, ask the right questions and conceive whole new visions from abundant data and information. This is where our industry must pick up a focus as big data’s intrinsic value declines. Specifically: We need to cultivate real seat-of-the-pants critical thinking around micro-employer data and macro-industry data. We need thinkers who will ask incendiary, never-before-imagined questions and propose changes and interventions that will reinvent how any employer’s WC program might be constructed and operated and how vendors will provide action and service. While vast, yet soon-to-be-cheap, data points will still garner some valid predictions, monetizing the employer’s change proposition and perhaps having a stake in the outcome will be where the future profit lies. Not just any claims expert can provide value at this needed level, as most in today’s world only know templates and best-practice concepts. Very few have skill in ground-up, project-based problem solving. The next wave of industry smart money must seek out and hire a new army of solution-prone human capital. Our industry must admit that predictive modeling is but an early step toward other means of value beyond just the current fleeting ability to increase underwriting profit or fast-track a claim process. The ancient industry construct that silos underwriting, sales and claims needs a re-assessment of where priority human capital investment lies and of how cross-skills must work together. See Also: The Science (and Art) of Data, part 1 Perhaps current position value will flip-flop… the soon to be data-rich yet bulk-automated underwriting process might become an offshore, outsourced common function while the adjuster will emerge as a future kingpin in protecting profitability and holding the highest salaried function – abundant with talent and intuition while provided ample time to ask the right questions employer by employer and claim by claim. I welcome any entity that wants to explore and build the next value-wave on the downside of big data to please contact me.

Barry Thompson

Profile picture for user BarryThompson

Barry Thompson

Barry Thompson is a 35-year-plus industry veteran. He founded Risk Acuity in 2002 as an independent consultancy focused on workers’ compensation. His expert perspective transcends status quo to build highly effective employer-centered programs.


Read More