by Joanne Basilio –
In the rapidly evolving digital world, disruption through technological innovation has become essential in maintaining a competitive advantage in the marketplace. Insurance companies try to be innovative but tend to follow historically proven solutions instead of boldly committing to culture-changing innovation. This trend leads companies unintentionally focusing on just improving what is in place instead of actively re-shaping their organization to face the challenges of the future.
With an overwhelming number of technological innovations to choose from, organizations can quickly drown in the myriad of options available while chasing the elusive “panacea” solution. When identifying the foundational elements, one of the focus should be on harnessing data to provide strategic insights into strategic changes. We have found that a high impact solution to help insurance companies harness the treasure trove of data that is available in house, third parties, and on social media is augmented analytics.
Augmented analytics uses machine learning and natural language processing to automate the process of data preparation, simplify complex data into a form that is significantly easier to interpret, and facilitate insight discovery and distribution. While augmented analytics and Artificial Intelligence share the same underlying technologies, augmented analytics differs in focus and application. Artificial intelligence (AI) builds systems that run without human intervention. On the other hand, augmented analytics is used in the creation of applications that enhance human performance. Rather than replacing human intelligence, it is meant to supplement the human’s natural ability and curiosity to recognize patterns and draw conclusions. Instead of turning the decisions over to a machine, humans remain at the center of the decision-making process.
Bring Together Data from Disparate Sources
Within insurance companies, we typically find siloed data sources that range from the 30-year-old legacy mainframe policy administration system to the 10-year-old claims system running on Oracle to the recently developed mobile application, all storing the definition of the customer differently. How does an insurance company glean insights across multiple platforms without spending millions to bring together data from various disparate sources and deliver a comprehensive view of their business without losing the essence of the data?
Augmented analytics platforms present a scalable solution to combine various sources such as databases, cloud services, enterprise applications, and even files from shared folders. Market available tools extend capabilities to collect and perform data analysis beyond the internal systems to include third-party data such as market trends, social media, secondary research, government data, and credit data.
External data can not only augment internal data but could be used to validate internal data findings.
Automate and Enrich Data Preparation
In traditional business intelligence efforts, raw data cannot be used as-is and requires standardization, cleansing, and structuring. The data preparation process typically entails a painstakingly manual process that takes significant effort and time, is error-prone, and produces limited re-use opportunities. Then, injecting humans in the process to review the data can introduce unintentional biases due to what is perceived to be relevant and impact the outcome.
In augmented analytics platforms, the data preparation process can be automated, drastically improving speed and accuracy. The ease of toll usage frees up data analyst time, which may be better used to tackle more complex algorithms, derive insights, and propose recommendations.
With augmented analytics, correlating relevant data, searching for patterns, and selecting appropriate models are done automatically. Compared to traditional business intelligence efforts, augmented analytics allows for faster processing of more significant amounts of data resulting in higher accuracy and broader coverage of permutations and combinations. Using automated data collection and preparation, the data sets are more precise and allow for a more distilled view of the data.
Collaborate Effectively with Data Literacy
Augmented analytics employs natural language processing (NLP) to allow for programming in English, allowing for effortlessness interaction between man and data. Using NLP, the need for the nearly impossible to find deep technical experience with business acumen can be shifted to a technical business resource.
This allows the end-user of the data to be closer to the source and reduce the number of hand-offs, delays to due misunderstandings and incorrect development. More importantly, it makes data readily accessible to enterprise users and allowing them to use valuable information and insights into decision-making and business planning.
Instead of using a technical diagram that requires a computer science degree, augmented analysis has been built to create complex visualizations of data to quickly recognize patterns and develop predictions. No longer are users limited by pre-built limited customizable reports and dashboards. Now users can design and program visual representations of the data tailored to each user’s style and understanding. This ability to redefine on the fly by the business resources enables more organic and enriched ideas to be built by a variety of subject matter experts in a collaborative manner.
Focus on Outcome-based Use Cases
With increased computing power at an ever-decreasing cost, augmented analytics will enable insurers to tap into external data sources like the internet of things (IoT) devices. Insurers may use the data from smartwatches and fitness trackers to provide input on customer lifestyle in predicting and calculating risks for life insurance. For example, policyholders could receive a discount if their wearable documents the amount of exercise.
Real-time connectivity will allow insurers to meet the demand for microinsurance and gig insurance as the economy evolves into its next phase. Through the use of augmented analytics, the insurance companies can build models that take into account multiple sets of factors such as the age of the car and driving history. Softer attributes such as frequency of social media posting, frequency of trips, time of the day, etc. can be taken into account in factoring the risk. Building these factors into a model, the insurance companies will be able to provide a streamlined process and more accurately price the risk.
As the world continues to become more complex, so does the new risks the market demands to be covered. As the risk and new products become more challenging to evaluate, so must the sources for assessing the risk. For example, a new hot product is cyber-insurance, which requires a new set of technical skills to determine. This risk does not rely upon the historical factors of building location but now must analyze unfamiliar sources like threat databases, dark web catalogs, and network security logs. As this is a new product, the models have not been perfected and will frequently change as new analytics become available. This is an area where augmented analytics will excel as it will be able to pull together all of the various data sources into a visual representation enabling the underwriter to define and price the risk accurately.
Not all analytics are about pricing or evaluating risk; it is possible to use augmented analytics to develop sentiment analysis. In this area, the analytics looks at the reputation and perception of the company from the customer’s view through reviews of the company in social media and other interactions. Through the analysis, it is possible to evaluate millions of posts to determine if the writing is positive, negative, or neutral. With the visualization of augmented analytics, it is possible to sort through social media data to monitor brand/product reputation and customer experiences. Also, it is possible to search through social media to distill ideas and concepts to improve existing products, build more meaningful customer relationships, and identify new product offerings.
Deliver Value with Augmented Analytics
Gaining the powerful insights of the augmentation analytic toolset does not rely upon technology alone. Instead, it is a well-thought-out strategy focusing the effort on achievable goals before broadening the horizon.
Since augmented analytics may be applied to a variety of use cases such as customer retention, user experience, sales, and productivity, organizations must start with those they see will deliver the most value in the current state of their business. They should articulate what measurable improvements in critical metrics each use case will contribute. We are seeing in our customer base the focus has been on customer retention and risk reduction as with everyone being remote, they have the time to shop around.
To build a robust analytics group, the team must be cross-functional with resources from sales, policy, billing, and claims being paired with IT personnel to enable the vision. The combined group will provide a well-rounded perspective on where to focus the efforts and what data sources to access. This will not only define a holistic view but build the all important change champions in the various organizations as this technology can have a bit of an acceptance curve to overcome.
As with any modernization efforts, it is a staged journey to the end goal. The plans must be laid out for today and tomorrow but must be flexible enough to adjust to the findings. No matter the end goal, the journey must account for the changes that are coming.
The changes include efforts such as modernization of the IT architecture, refreshing of databases, but more importantly, identifying the impact on the organization. Personnel in the organization will view this as a threat to the way they do business and must be reassured that their skills are not being replaced but rather augmented with a new digital co-worker. Now, staff can quickly access, review, and identify trends in mountains of data that they will need to communicate and act upon effectively. The team must evolve to focus on higher-value activities of defining the next direction instead of trying to keep up with the current operation.
Just as important as having a good road map is selecting the right partner to take the journey. RCG brings a depth of experience in the analytical space from machine learning, to artificial intelligence to analytics. RCG does not look at projects as a singular transactional event but rather as a step along the journey of digital acceleration.