Data has the ability to transform an insurance organization through meaningful insights. Sometimes these words, though, don’t do justice to how much data can teach us and lead us toward greater profitability. Let’s think of data in terms of “talent” for a moment.
Imagine that you have been placed in charge of hiring someone who will lead your organization into the future with a whole new approach. Your supervisor gave you just one instruction. “This person must be a recent college graduate.”
It sounds preposterous. What recent student could know enough to lead the organization into the future? Yet, you found two excellent candidates.
- Candidate #1 graduated in four years, attended all of the right classes, read all of the required texts and graduated at the top of their class.
- Candidate #2 read all of the required curriculum, attended all of the right classes, but took 8 years to graduate, taking their time to complete 8 internships at 8 different insurance companies. At each successive internship, they were able to apply what they had learned in their classes and what they had learned at the previous internships to formulate an entirely new way of understanding risk.
Which candidate will carry you into the future?
Data that’s smart enough to teach and lead
Insurers are hiring one of these two candidates every day when they add data streams to their processes. Some data is fresh out of school. It is valuable, but its value is limited by its lack of experience and testing. Other data not only has experience, but the use of machine learning has given it intuition. It is able to predict with great accuracy. It grows even more valuable as it is aligned to additional relevant data points.
This is why Majesco is rapidly growing its Majesco Data & Analytics practice. Insurers are now able to take advantage of one of the most highly-prized P&C data intelligence products available. They will be able to utilize Majesco’s experiential data, integrated with a robust array of third-party data streams and analytic tools. These products will be enriched through artificial intelligence and machine learning. They will be deeply integrated with Majesco’s core solutions. The promise for insurers is a data culture of constant learning that will improve profits and lower exposure.
Initial Focus: Data that can act as a foundation for P&C analytics
Insurers who want to capitalize on the value of property inspection data must begin somewhere. If they are looking for a foundational P&C data set upon which to build a predictive analytics system, Majesco has the right data to act as the foundation – loss control data. The loss control data includes 16 million property surveys completed by trained risk engineers in the field and has been QA tested. Those 16 million surveys contain over 200 million photos and those 200 million photos are all tagged to the risk. It is the perfect (valuable and massive) data set upon which to applying machine learning.
Here is how the machine learning adds value to the data. The risk characteristics contained within the database allow Majesco to create models that can determine how much risk there is for almost any specific given property. For example, it allows you to take a picture of an electrical panel, a hot water tank, a roof, a backyard, and submit that to the software. The software will automatically analyze those images against the other 200 million photos Majesco has in its data, and it looks for matches that will provide insight. This is a key component in the automated decisions that carriers need to achieve better operational results.
It is as if insurers can now “look over the shoulder” of thousands of trained professionals inspecting 16 million different properties and gain the results of all that experience. How might an insurer improve its claims results with this kind of learning and experience?
Data that improves eye-sight as it improves insight
Insurers have every reason to love the insights derived from machine learning on 200 million images. It is an opportunity for increased automated decisions because predictive analytics can help identify known risks. The insights also work, however, to uncover unknown risks in a very practical way.
Right now, less than 10% of properties receive a physical inspection. Insurers would prefer, though to look at 100% of properties. There just aren’t enough inspectors to take on 100% of policies. If there were enough inspectors, there would still be too few underwriters to view the photos from the physical inspections.
If an insurer can start moving the process of taking photos of relevant property points to the policyholder or to the agent, they can look at more homes and more businesses, engage the customer, and relieve the burden of travel, time, and cost from using actual inspectors. Predictive analytics engines can then review what’s in the photo. Machine learning will automatically categorize the underwriting risk without an underwriter having to look at the actual item.
For insurers, this is like adding feet on the ground and eyes on the data – scaling the risk assessment and underwriting process, without adding any additional resources.
Data sets that continue to learn from other data streams
Everything we’ve mentioned so far is currently being done by Majesco client companies. Many of them are using Majesco Loss Control 360 to assess property and Majesco Property Intelligence model for property data insights and they are making informed decisions based the risk knowledge contained in the vast loss control data store. Very soon, with Majesco’s claims data, we are going to be able to create connections between claims and loss control surveys that will make the data even smarter. When we apply machine learning we will begin to see the opportunities for modeling that will make organizations even smarter than they are today. This will be a unique Majesco offering that will be difficult for anyone to duplicate.
Data’s long-term value for the organization
All of the above is exciting. It is a crucial part of how the insurance industry is going to move forward in a cycle of constant, data-fueled improvement. Majesco is creating a number of data and analytics products like Majesco Property Intelligence that are pre-integrated with Majesco Policy for P&C, many which will be available through the Majesco Digital1st® EcoExchange.
The long-term data strategy, however, will have a broader appeal to operations, underwriting and finance by impacting three key areas:
- Smarter workflow and straight-through processing
- Improved underwriting and pricing accuracy
- Identifying appropriate coverages
Majesco now has the partnering power and the expertise to create a large contributory dataset that can be augmented with other data sets and technologies to do very significant machine learning and analytics.
If you are familiar with commercial insurance, you know how complex the forms can be. Straight-through processing is that catch-all term for anything that will improve workflow, from form field filling to the automated gathering of risk data. Machine learning and analytics have the potential to make every step faster, easier and more streamlined. Every step that is saved reduces costs.
Improved underwriting and pricing accuracy
Besides new business processing, tomorrow’s enhanced data sets will help insurers to automate the underwriting process and to more accurately underwrite their policies. They will help insurers to identify appropriate pricing and to only select risks that fit their preferred risk profiles. The result will be reduced losses and improved loss ratios through lower claims.
Identifying appropriate coverages
Premium leakage is a big deal. Companies are missing lines of coverage because they don’t have a complete picture of what’s appropriate for all of the properties they insure. When insurers decide to go after all of the data that they can pull together, AI is going to help them identify missing lines, potential fraud and more. They will be able to go one step further in claims with automated adjudication that takes into account all of the metrics of the claim: the amount, past claims history, etc. The analytics will have eyes on policy overactions. Using AI/ML, insurers will be adding a valuable tool to provide protection and consistency to their own internal policies.
Majesco Data and Analytics: Products, strategies, services.
Step one in Majesco’s new data initiative includes the first release of Property Intelligence powered by Majesco Analytics. More products and services are on the horizon! Is it time that you begin adding a data strategy to your goals for automation, machine learning and bottom-line impact? Contact us today to get started.
By Patrick Davis, SVP of Data & Analytics, Majesco