Marketing & Distribution | Read Time: 7 Minutes

Data Insights and the Future of Insurance

July 9, 2020

Brian Wood By: Brian Wood

I’ve spent the better half of 20 years working to develop technologies and data insights with the goal of transforming the insurance industry. Starting out first as an agent, I continued with multiple insurtechs as well as some of the largest players in the space. I’d like to share a few key lessons I’ve learned along the way as well as how data insights are changing our industry right now.

We can’t talk about data without touching on the digital revolution of insurance.

There are four drivers of change across the ecosystem:

  • Consolidation is making agencies larger and more complex
  • Commoditization is driving the need for specialization and higher levels of productivity
  • Customer expectations require new, simpler user experiences
  • New competition is putting a lot of pressure on insurers and brokers to become smarter through data and analytics

While COVID-19 will prove to be one of the most challenging times for most of us, both personally and professionally, one thing is clear: to succeed in today’s environment, the digitization of insurance has officially moved from a competitive advantage to table stakes.

With that in mind, I want to discuss a couple of industry problems that we are tackling here at IVANS:

  1. How to identify markets quickly and successfully
  2. How to translate data into revenue growth, increased retention and reduced costs

Solving the Problem of Identifying Markets

Even within the last few years, a lot of commercial insurance distribution was a combination of insurer sales reps leaving paper brochures, agents going to multiple insurers for appetite guidance, and agencies managing spreadsheets that were out of date or inaccurate.

In discussions with insurers, many insisted that their agents knew their appetite despite an industry average quote rate of 40%. This rate means that 60% of submissions were declined. We had two key champions that drove us forward. The first being the data, which revealed a huge problem. The second being the individuals within agencies – the CSRs, account managers and the producers, especially ones without 20-30 years of experience – who told us that there has to be a better way.

To address this problem, IVANS began working with insurance companies to digitize their appetites, many of which today are delivered via APIs. We then built integrations into nine agency management systems, which host roughly 60-70% of the commercial business in the U.S.

Five years ago, agents would spend between 20 minutes to sometimes days to identify the markets they should be sending submissions to, and now it happens within seconds. Many of our insurer partners are even utilizing our APIs to drive their appetite guidance within their own portals. When you look at the numbers, it turns out this was quite a recipe for success.

Market Appetite has been growing rapidly:

  • Monthly Requests: 6,178,685
  • Agents: 124,601
  • Agencies: 14,441
  • Recommendations: 54,636,578
  • New Commercial Policies Placed: 141,781
  • New Commercial Premium Placed: $702,380,600

These are strong metrics to prove the success of IVANS approach: we digitized information, developed needed insights and distributed those insights. All of that is a great start, but we all know insurance is a relationship-based business. If people are not in the forefront of your equation, even as it relates to data, platforms, APIs, and all of the other terms we like to use, you will miss the mark.

It comes down to what users believe about the technology supporting them. This includes your own applications, so ask yourself:

  • Are your users excited about the problem you are solving?
  • Are you seeing strong adoption?
  • Is the value of the insight strong enough that it creates champions that then evangelize your product to others?

By taking a relationship approach, we have been able to create a win/win/win solution. More business for insurers, a much more efficient and scalable process for agencies, and better outcomes for the end insureds.

Solving the Problem of Turning Data into Actionable Business Insights

In addition to leveraging data for identifying market appetite, data-driven insights are playing a vastly increased role in major organizational decisions, driving front office sales strategies and the enablement of back-office goals. But how do you use them in your own organization?

One trend you are likely seeing is adding external insights to provide a broader picture of the problem you are trying to solve. This is to say that often data and analytics only reflect what a company can discern based on its own data. But it is data insights related to your performance versus your peers that end up being the most actionable. Additionally, having a large and well-structured data set is critical otherwise, as you divide data by industry, geo, dimensions of size and other influential attributes, you very rapidly move from cake to crumbs. 

So let’s say you have a solid data set; what problems are we trying to solve?  A key attribute of a strong data insight is its action-oriented nature. Very often, analytics and data provide insights that are interesting but translating it into revenue growth, increased retention or reduced costs can be very challenging. 

One of the biggest challenges IVANS had to overcome was ensuring that any insight we produce has a statistically significant amount of data behind it and is accurately correlated. This is why the traditional approach of slicing and dicing data, using artificial and rigid size bands, and then averaging the results to get pricing insights is fraught with inaccuracies.

Within insurance, we often run into challenges within the data. Missing key data elements such as industry classification and similar, can make even large data sets next to worthless. This is where machine learning comes in. We now use machine learning to help us identify and/or confirm industry classifications within our data set. Once identified, we are able to use other machine learning models and the experiences therein to continuously learn and improve our results.

One of the largest challenges our agency and insurer partners shared with us is the lack of pricing insights for commercial products. Agency producers need this intel to make renewal marketing decisions and to support the ever-increasing demand for data from their customers. Insurers need this information to make smarter decisions about their pricing. Today, it is limited to publicly filed rates, analyzing data within their four walls and information coming from the field.

Pricing insights are highly limited but incredibly important for both insurers and agencies. As an insurer, how do you understand your won/loss ratio results as it relates to competitiveness? How do you know how your pricing compares? To address questions like these, IVANS began developing machine learning (ML) algorithms and applied them to our data set. It took many months of comparing the performance of competing models, training those models and ultimately identifying the various data attributes across size, industry, geography and LOB that are most correlated with variations in price. For BOP alone, we use 64 different ML models.

We then worked with agencies and insurers alike to understand how they would like to ingest these insights. Many want an annual or more frequent evaluation of their book while others want to focus on specific areas they see as either being a problem or an opportunity. IVANS is working with several of them to incorporate the insights within their quoting processes or within their CRMs. To support the use cases, we developed two ways of accessing benchmarks: either as a data service to augment and work inside their own business intelligence tools or via API.

By using IVANS’ data set coupled with machine learning, we are able to provide a more concise way to help determine if your pricing is in the ballpark. With this new lens, you can go further, gaining a host of insights such as reviewing particular segments as well as identifying potential strategies based on this data.

What’s Next? Upcoming Data Insights

IVANS has identified benchmarking as a huge need and will be expanding to other commercial lines as well as personal lines:

  • Premium Benchmarks for Workers Compensation, General Liability, Property, E&O, D&O, and Personal Lines
  • Premium Renewal Change for states and industry classes
  • Policy Coverage Limits and Deductibles
  • Line of Business Recommendations and Account Rounding
  • Claims and Losses

We are also making significant investments on coverage benchmarks, deep agency analytics that drive sales and operational efficiencies, as well as eventually providing claims guidance.

If you are interested in using data-driven insights to optimize your pricing and marketing strategies, learn more about the IVANS Benchmarks application.

Brian Wood

Brian Wood

Brian Wood, Former Vice President of Data Products Group for IVANS, develops innovative and data-driven applications, tools and services for the insurance industry. Prior to this position, he co-founded EvoSure, which was acquired by Applied in Sept 2015. Brian spent 10 years with Marsh & McLennan as Senior Vice President of Strategy and Business Analysis where he developed global insurance distribution platforms. Brian began his career in insurance building the first company to bind insurance online, (a Trilogy company) which was acquired by Marsh & McLennan in 2001.