How New York Life raised its customer service game

February 23, 2023 6:30 AM

New York Life Insurance Company's New York headquarters building.

Digital Insurance spoke with Alex Cook, head of strategic capabilities at New York Life, on the life insurance company's initiatives to modernize the carrier's operations. One of these products is GuideMe, a platform for working with insurance customers to collect facts and to set goals for their life insurance coverage. This unit of the company is also responsible for MyNYL, which simplifies and streamlines ongoing customer interaction such as changing coverage, paying premiums and making claims. Onboarding new life policyholders required New York Life to address a complicated health industry infrastructure when collecting health risk data, as Cook recalls.

What do the electronic health records used to weigh life insurance risk look like?

Alex Cook, head of strategic capabilities, New York Life Ventures

The vast majority of that information comes from the providers – doctors or hospitals – having their own electronic health record platforms. There's a handful of companies that have a large market share like Epic, Cerner, and Veradigm. These platforms are used in-house by providers to help manage all of the health records of patients. But a request from a life insurance company always comes with an approval from the applicant to obtain the health record data. Typically, the healthcare provider would print out a record from their system and then fax it into the life insurer. That's a very cumbersome process. It can take weeks to get the record and when you get it the quality is not always that great. It's in image form. It's not so easy to extract the information.

New York Life was really the first in the industry to reach out to some of the electronic health record platforms and convince them of an important use case for life insurance. Epic was the first company that we really convinced of this opportunity and they created a platform for use by life insurers that now enables all the carriers that use that system to get that data electronically. It's now much easier to review and summarize the information and even extract some of the data for use in rules and models.

[Editor's note: Since this interview was conducted in fall 2022, Cook added that NYL has also "partnered with state-sponsored healthcare information exchanges, and with Human API to access patient portal and other data."]

How is New York Life using AI and data science?

Data science is used a lot in digital marketing, and we definitely engage in that. We also have a lot of focus in the underwriting domain. Some models predict where we can, for a given applicant, bypass the need for blood and urine. A lot of carriers have increasingly focused on that area because, along with electronic health records and attending physician statements, that's one long pole in the tent in underwriting.

We have models that look at information that's more readily available and they determine whether there's a need to collect fluids to make a good underwriting decision. Many times, that lets us bypass the need for fluids. We also have predictive models that directly predict mortality.

That's something that we're continuing to enhance and roll out to a larger proportion of our policies. We're starting with the somewhat simpler, smaller policies and then expanding from there. It's a way of using data science and AI to make the underwriting decision about risk class rather than have an underwriter go through that data.

The AI that we develop here is not black box. It is a structured, supervised learning approach that we adopt because you really do need to understand the reasons why you are assigning somebody to a given risk class for underwriting. We want to make sure that it's grounded in good understanding of the actual mortality implications of the data that we're seeing. A staff of doctors has co-developed the models with us. It's important that we develop that in a way that's clear and transparent, why is the decision what it is, what it's based on, and it's grounded in good medical science and actuarial science.

How does this view of mortality models shake up life insurance illustrations?

It doesn't affect the illustration so much. Risk class does affect the cost of insurance that's in the illustration. The illustration also reflects the forward-looking assumptions around what the crediting rates might be: If it's whole life, what the dividend might be over time, or if it's a variable universal life policy, or what the assumptions are about returns on the investment sleeves of the policy.

There may be some differences in risk class assignment, whether you have a model making that determination or whether you are using your standard underwriting manual to make that determination. But those variances are typically not significant.

GuideMe collects in a robust way all of the financial information about the client, what their goals are. Then we can illustrate graphically the combination of products that better meet the needs of the client, and they can understand what that looks like and they can easily dial up or dial down coverage. The illustrations, which are a regulatory requirement, are still run in the background. Once the client and the agent agree on the package the client wants, it's there and available. They get that information and can understand it much more easily in a much more graphic way and a much faster way.

What’s the state of awareness of the need for life insurance?

It is critical because life insurance does play an important role in stabilizing families and communities. Over decades, we've had a long term secular decline in the proportion of Americans who have any life insurance coverage, let alone sufficient coverage to meet their needs effectively. So there's a very big gap in life insurance ownership today.

There are a couple of factors driving that. One is over time there's been introduction of a lot more financial products. Mutual funds, ETFs, a lot of focus on other financial instruments. And so there's a sort of mind share element there that makes it complicated. The other key reason is a very long term secular decline in the number of life insurance agents. When you're talking about a complex financial picture, talking to somebody – an agent or advisor – is critical, not just to understand the space and what you might need to serve your needs effectively, but to enable you to take action.

With a lot of the insurtechs and their entrance into the space several years ago – Ladder, Ethos or Policygenius, et cetera -- the results are pretty clear at this point. They grew pretty well from a zero base, but they're still at a point where the volume that they're selling is well below what was expected when they first came on the scene. They have all realized the need to include agents and advisors in their business model. Some of them have been hiring their own agents. Some of them have pivoted to being more of a platform provider for independent agents to use for their business.

They've been also looking for ways to engage with each other and cross leverage. That all points to the ongoing need for supporting people with agents and advisors.

A lot of the digital focus has been to make the experiences more seamless and easier and faster so that you're reducing those barriers for people to buy life insurance. We need to continue that, but we also need to continue to raise the awareness of how important it is to have that coverage for your family. It's important that we continue to raise the importance of life insurance and other insurance products, and do what we can to support it by making the product more easily bought and better supported by more agents and advisors.