AI can power faster payouts by insurance companies

The whole business of financial services in essence is a pay in-payout process. In banking it’s called deposits and withdrawals, in mutual funds it’s purchase and redemption and in Insurance it’s policy issue and claims. 

A claim is nothing but a Payout from customers’ collective fund which insurance companies hold in trust on behalf of all policy holders.

Since its customers’ own money, it’s important to note, settling claims fast and fair is the primary responsibility of insurance companies and it is here, most of the user experience gets impacted. 

Can AI help? 

Before we get into that question, let’s accept the onus for fast and fair claim settlement rests on both the parties. Customers should claim only what’s fair and companies should pay without a hassle what’s due.

Determining what’s due is not a rocket science. Notwithstanding the army of underwriters, claim managers, compliance and claims’ committee members working hard, there is a  tendency to complicate things in what should, otherwise, be a straight process.

It’s actually a simple 3-step journey:

1. Check how much a customer is eligible to claim under the policy?

2. Check the evidence submitted validating the claimed amount is genuine and not doctored

3. Check the person claiming, is the one actually entitled to receive the money under the policy.

We may do any amount of hair splitting to add/modify these 3 fundamental checkpoints in claim settlement but if you are through with these checkpoints, claims are payable. Period. 

Now, why should it take an army of people to process 3 seemingly simple questions? Let’s examine these questions a bit closely. 

Checkpoint 3: It’s essentially a KYC step. Digitisation has solved this step across banking, travel etc to generate responses in seconds. 

Checkpoint 2:  If insurance companies do not obfuscate and take refuge under small print, this step too can be solved even faster. It’s essentially  a Go /No Go toggle. It should never take more than a few seconds to pass through this checkpoint even manually. But unfortunately, insurance businesses needlessly complicate things. Sum insured (the total amount payable as claim),  for example in a health policy, never is a straight forward ceiling. It’s divided into limits, sub limits by category of benefit (say room rent or ICU charges)  and timing (per day limit etc).

The question to ask is, is it required to be that way? The answer is No. Making life difficult for customers is never a good business tactic.

But then why do insurance companies tie themselves in knots and make customers pass through a minefield of seemingly unintelligible terms and conditions?

It’s simply lack of data, intent to be on the customer side or enabling tools that prevent insurance businesses to be simple and easy. Now that AI has arrived, we just need the right data and systems’ framework to sort the step 2 in seconds.

Step 3 is a bit more complicated. Its validation requires objective and subjective scrutiny. Whether the claimant is acting genuine or fraudulent, needs deeper analytics.

But, even before AI arrived, didn’t credit card companies and online banking solve this problem? What’s a suspicious transaction and what’s not, can get sifted in seconds by using fraud analytics. But then why do insurance companies fail to process claims instantly  while credit card companies across the world, every day, settle millions of transactions in super quick time? 

The answer is the nature of the business model. Credit card companies have a vested interest in approving credit instantly. They don’t exist if they don’t. On the other hand, insurance companies make money by rejecting/delaying claims. They lose money if they settle even genuine  claims instantly. For insurance companies every rupee held longer and not paid instantly, earns them investment income and adds to their bottom line. In fact, some of the companies survive only by rejecting/delaying claims.

How long can they cling to this business model of making more money by caring little about customer delight?

Well, they got away with their faulty business model for long. But with AI, perhaps they can’t any more provide excuses. They must now invest in data analytics and digital systems that enable AI. If not, they may follow the way Nokia or Kodak which got drowned in the digital wave. 

 (The writer is Founder CEO of Upsure, an InsureTech solutions provider)