Life Insurance Is a Service, Not a Product Line

Atidot’s name, Dror Katzav explained, is the Hebrew term for picking the right alternative among multiple futures, a fitting name for a company doing predictive analytics over the long horizons that life insurance involves.

About Dror Katzav

Dror Katzav is co-founder and CEO of Atidot, which he founded in 2016 after about 11 years in Israeli intelligence as a software developer, team leader and product manager. He teamed up with a former technologist from his unit to bring data solutions to an industry that could benefit from them, moved to the US about two years before the recording, and has raised about $9 million. Atidot helps life insurers understand how policyholders use their policies and how policies behave over time, in order to improve the profitability of the book.

Why Life Insurers Care About Lifetime Value

For years, Dror said, life insurers focused on getting customers on board, and low interest rates and the economy covered for whatever else happened. After 2008, interest rates stayed low and customer expectations changed, so the biggest asset, the existing customer base, wasn’t being managed as it could be. Understanding the policy and its life enables better alignment of product, through upsell, cross-sell and retention, and rethinking of pricing, distribution and asset-liability matching. That’s a data science exercise now that insurers can bring in far more data.

The shift is from a product mindset to a consumer one. If a customer buys term life and eight years later converts to whole life or universal life, the lifetime relationship is worth far more than the small term policy, and that changes how much an insurer should spend on acquiring, servicing and retaining them. Dror used a manufacturing analogy: insurers think of themselves like Target, producing policies to ship out through distribution partners, when what they actually sell is a service, and consumers have been trained by Amazon, Spotify and Netflix to expect more.

Selling to Every Layer

Most of the people who see this value are in the C-suite, but Dror’s point is that a software product can reach every layer. Formerly it was a consulting engagement from a firm like McKinsey, BCG or Milliman, which seldom turned into action. With software, a chief distribution officer gets a distribution-strategy dashboard with benchmarks and gaps, the middle layer can slice the data, and production-floor staff can get real-time recommendations on specific policy orders, including leads pushed into a CRM.

Life Versus P&C

In P&C, the contract is renewed every year, which gives a touchpoint to rethink pricing, risk and engagement. In life, you underwrite a customer once and are stuck with them for life, sometimes a blessing and sometimes a curse. The data is different, with longer horizons and fewer touchpoints, and so is the modeling. Turning a life insurer into a data-driven business is a different challenge.

Why Life, From Talpiot

Dror explained the culture that led him to the area. His unit would receive the government’s toughest, least explainable objectives each year, and product managers had to work out how to answer them with data, technology and operations. They looked for the questions everyone said no one knew how to answer. P&C had many companies working on it, but in life insurance, he said, nobody touches it. They preferred the unbeaten path where competition would be light. Dror also described the Talpiot program Gilad and he share: a bachelor’s degree in physics and math combined with basic training across the army’s branches, followed by nine to ten years as an R&D officer, with both the practical and academic background.

Benchmarking Against Peers: The Electoral Map

Atidot works extensively with Pacific Life and Guardian Life, among others, and with iPipeline, which Dror described as the most dominant platform at the point of sale in the US. The relationship gives Atidot access to the data of policies being purchased and the ability to run models and show carriers not only their performance but their peers’, anonymized and with the privacy of brokers and carriers preserved.

His example: a carrier grew term sales five percent and thought that was good, while everyone working with the same distributors, segments and products grew 15 percent, so the carrier left 10 percent on the table, which can be quantified in premium, perhaps $5 million to $30 million. Atidot breaks it down by distributor like an electoral map:

• Red states: distributors that are small for the carrier and not growing, not worth the attention.
• Blue states: big and growing, where the carrier has already won.
• Swing states: big but declining (someone is sneaking in), or small but growing, which are where to put effort.

A second layer looks at operations: where the process is leaking, for example too many medical requirements or too long to deliver a policy. It’s a joint product that can be bought through Atidot or an iPipeline representative, and Dror said it was “flying off the shelves.” At Pacific Life, the first customer, a weekly report goes to the chief distribution, operations and product officers, the CFO and about 60 other people.

Taking Ownership of the Next Mile: Guardian

At Guardian, the relationship moved from insight to operating on the insight. Atidot is taking ownership of a block of business where the selling agent is no longer part of the business, so policies are orphaned, reassigned or unassigned and receive poor service. The models suggest that better service could double the premium, a 2x gap between the part of the book that gets service and the part that doesn’t. The aim is to bridge the insight to the home office, distributors and agents on the front lines, in a hybrid digital and offline model.

What’s Next: The Second Policy

Dror sees the opportunity in taking a holistic view. Most customers have no tool to find out whether their policy is good for them beyond reading a 600-page PDF, and with an aging agent population, a hybrid solution is needed. Other life insurtechs go after the first policy and the millennial buyer, while the overlooked segment is the customer ten years later, already paying for 10, 15 or 20 years.

Gilad asked whether insurtech MGAs and new intermediaries could be customers. Dror said many will only meet these challenges in years five to eight, since early on they sell cheap term online and see churn before purchase. Post-level term, when a 10- or 20-year policy ends and the insurer must offer a conversion, is the hard part: raise the price and you keep bad risk, don’t raise it and you’re stuck with bad pricing. Younger insurtechs haven’t reached that stage, he said.

On the question of B2B versus B2C, Dror said they are materially different professions. Lemonade, Hippo and Next built brands and consumer products, while Atidot is focused on data, processes and analytics for carriers, working with existing portfolios, with hundreds of millions of dollars of premium. He believes life insurance is the close follower of P&C, which trails banking, so insurtech in life is next.

Advice: Wear Shoes

For his recommendation, Dror chose a small work-from-home ritual: he puts on socks and shoes every day, even though the office is ten steps from the bedroom, and takes them off at the end of the day. It helps him separate when he is working from when he is not. Gilad added that most of his sleeping shirts are startup swag.

Key Takeaways

• Life insurers have long managed products, not customers; a lifetime-value view changes how much to spend on acquisition, service and retention.
• In life, you are stuck with the customer for life and data comes with long horizons and few touchpoints, unlike annual P&C renewals.
• Atidot’s access to point-of-sale data through iPipeline lets it benchmark each carrier against peers, quantify premium left on the table and map distributors like an electoral map.
• At Pacific Life, a weekly report goes to about 60 executives across the organization.
• At Guardian, Atidot is taking over service for orphaned and unassigned policies, where better service could double premium.
• The overlooked segment is the policyholder ten or more years in; the hardest data challenges, like post-level term, appear around years five to eight.