(American Gothic, Grant Wood (1930), Art Institute of Chicago.)
Many (re)insurers refer to themselves today as data-driven organizations, but the reality is that most are not. This perpetuates much of the industry’s traditional legacy mentality. In a true data-driven organization, data flows up, down and even side-to-side. Data empowers every employee, from the most junior level all the way to senior executives, to make decisions, and to communicate the rationale behind their decisions and motivate action.
For an industry that relies heavily on accurate information, most (re)insurance companies today don’t have enterprise-wide big data strategies. Compared to other data-intensive industries, which aren’t encumbered by the legacy issues that continue to hamper (re)insurers across the value chain, that’s something that needs to be addressed since business intelligence and analytics are now top priorities in the industry, according to many studies.
It’s true that, historically, the majority of those in the industry have generated significant value from data—even in its rawest form—to gain insights into their operations, e.g. to assess risk, develop pricing models, streamline claims, enhance channel distribution, etc. However, in many cases the value of those operations for which big data’s promise was intended oftentimes is marginalized due to the lack of use of technologies designed to realize its potential.
For example: Loading underwriting source data into a traditionally structured database and relying on a few stored procedures that a programmer wrote years ago to “automagically” render high definition insights today is foolhardy. In cases like that, especially with the vast array of underwriting data available today, underwriters are lucky enough to pull a few key data points to effectively do their work with the hope of increasing underwriting profits and lowering loss ratios. “Big data” in that case like is relegated to self-defeating “small data.”
Unleashing Big Data’s Potential
Every worthwhile data strategy includes a component to harness the huge amounts of data the industry amasses every day. Companies that have decided to champion their data have bought into the concept of creating a data ecosystem designed to do at least three things:
- build upon their data foundation;
- create ways in which their data can easily be made accessible by the enterprise – not just in silos; and
- develop various means to enhance their data, both from a completeness and accuracy standpoint. This is about creating a single source of truth that enables everyone to work from the same place and use common processes.
As recently as five to ten years ago, harnessing data within the (re)insurance enterprise was much easier than it is today. However, by the same token, the benefits of using data at that time were much more limited in scope. The ability to harness big data in today’s insurance world involves leveraging combinations of more data sources than ever before – social media is one example – including new technologies that not only provide insurers access to data, but the ability to make sense of it and put it to work.
The vast majority of today’s industry data is unstructured (think: text, audio, video, adjuster notes, click streams and log files). Then there’s structured data, which typically comes in the form of peril, geospatial, and even traffic- and weather-related data, among other types. (Re)insurers must account for that unevenness when capturing and synthesizing those various data streams. Remember, big data analysis involves finding correlations and patterns that might not otherwise be observable, and that almost necessarily involves uses of data that were not anticipated at the time the data was gathered.
The pervasive problem, however, is that without much manual intervention and extraneous workarounds often found in companies’ operations, it’s difficult for traditional data technologies to unleash big data’s potential. The chief benefit newer technologies bring is their ability to handle data in any form without having to develop time consuming processes to (re)integrate data once it’s “cleansed.” That alone increases agility to respond to often-changing business demands. At the same time, it encourages innovation and cuts much of the costs associated with traditional relational data management platforms. As with most everything else, there is a balance that calls for employing relational and non-relational enterprise technologies, considering the opportunities offered by new technology paradigms, while continuing to utilize the old paradigm where appropriate.
The Data-Centric Insurer of Today
Today’s data-centric insurers must combat traditional means of digesting massive amounts of data by relying on a different class of technology platforms. Perhaps the most popular today of that class is the open source Hadoop ecosystem and its commercial variants. Hadoop is flexible enough to work with multiple data sources, either aggregating multiple sources of data in order to do large-scale processing, or even reading data from a single database to run processor-intensive machine learning jobs. Putting to work the power of data technology platforms that were were specifically architected to process massive amounts of data from which to gain actionable insights is no longer tomorrow’s hope, but today’s standard.
Much like many newer technology platforms today, the cloud should be seriously considered when developing and executing a big data strategy. In fact, an effective big data strategy embraces cloud computing in all its glory. Cloud computing enables burning-fast analytics computation at a fraction of the cost than when cloud computing became mainstream roughly ten years ago. The main reason that cloud computing is so complementary to big data processing is that it drives companies to acquire and store more data than ever and, in the process, creates more need for processing power to drive the so-called “vicious data circle.”
Deriving Value from Big Data
Big data’s value comes in many forms. For starters, if leveraged as intended, big data alone offers greater transparency to the enterprise. By simply making appropriate data more accessible to company stakeholders, companies stand a better chance to make more informed decisions, and in many cases even critical decision-making can even be done in real-time.
Secondly, as data transparency increases across the enterprise, with a sound big data strategy in place, information is put to use quickly and accurately, which enhances operational efficiency and opens the door to more free and innovative product development thinking. Obviously, offering competitive products and services is a significant market differentiator. For example, the insights big data bring in that regard help (re)insurers arrive at incredibly accurate scoring models and readily take into account customer behavior, which aid (re)insurers today to tailor individualized products and pinpoint market segmentation models based on multitudes of risk factors.
Gaining greater insights into customers via big data enables companies to better understand their needs and affords a clearer path to tailor products, develop services and even target customer segments.
Lastly, analytics designed to surface actionable insights while mitigating risk plays a crucial role in big data strategy. Companies that devote resources to gathering data and undertaking complex analysis do so because it’s in their interest to make more accurate decisions, which, if done as intended, has the likelihood to positively affect the bottom line.
It makes sense for (re)insurers to embrace big data; big data can play a significant role in a company’s future success if truly understood and employed throughout the enterprise. Along the way to embracing big data, companies must modify processes and adopt the technologies that best support how they ingest data in terms of type and amount, and even the speed at which data can be processed and consumed.
In the end, (re)insurers that excel at that and embrace a data-driven culture will intensify customer relationships, have the ability to more nimbly move into new markets, but perhaps most importantly, catch up to—or even surpass—more nimble companies already in the game and posing a disruptive threat to industry status quo. They will win the best clients, mitigate risk, increase customer loyalty and create more opportunities to develop, market, sell and cross-sell products and services. The rest in the pack will risk falling behind or ceding the most attractive customer relationships and emerging markets to others.