3 Learnings for Successful Commercial Underwriting
How to prevent fraud throughout the full policy lifecycle with data-driven real time decisions
About The Author
An insurance professional with 17 years experience working in claims, sales, distribution and broking. Having worked across the whole insurance lifecycle, Martyn brings that experience to bear, helping insurers use software to make better decisions in underwriting, claims and compliance.
Martyn Griffiths
Sales Manager UK | FRISS
When I started my career in insurance, I was in a local branch, dealing with claims, and the underwriters and claims team sat across from each other. We could ask each other questions across the filing bay down the middle of the room. The underwriters had an understanding of the businesses, trades and hazards in the local area, and if I had an issue with a claim and wanted to understand the intent of the underwriter when dealing with a claim, I just had to holler.
In today’s world, the claims and underwriting operations can be geographically separate and increasingly underwriting and claims decisions are automated. Commercial Insurance is also being disrupted. This provides a great opportunity to come away from the silo approach to fixing something in claims, or underwriting. The opportunity exists to deploy tools that help remove silos and foster greater insight and co-operation across functions.
The Benefits of Breaking Down Silos
There is a difference between KYC upper case and knowing your customer well lower case. If you have a clear, up-to-date view of individual risk and your portfolio, you will be able to underwrite with greater confidence in a market where exposures are high, that could avoid writing business that leads to a high-value claim. So how do we do this? There are three core building blocks:
Real time network analytics as new risks and claims come into the portfolio. What I am talking about is the ability to understand all the interconnections between all the subjects and objects in both claims and policies. To pull back the stakeholders behind the business and understand their involvement in other policies and claims. Was that a good experience? How does that information help us make a better decision, either in claims or underwriting?
The ability to understand what the business is doing in depth, does that fit your risk appetite, supplementing and verifying the standard question sets to gain more insight.
Constant monitoring has there been a change in the stakeholders of the business? What does that mean? What about changes in the activities of the business. A general builder who is suddenly going all out for roofing work mid-term.
“Real Time” vs Real Real-time
AI is a vital part of the conundrum. Network Analytics will provide insurers with more data in context to allow models to make better, more predictive and data-driven decisions. Whether this be about predicting a heightened propensity to claim when you are writing a risk, spotting a pattern in application fraud, or predicting the presence of fraud within a claim.
In itself, this is not something new. However, typically the sort of processes involved are batch submissions of data and rebuilding of static networks overnight or at other frequencies. Others offer the ability to query the static batch network in real time. But this is not actual real-time. The ability to understand the impact of a new business submission and its relationship with the portfolio in proper real time, is that it is showing the connections of that submission to a claim made a minute ago, or another business put on cover immediately prior. This is the opportunity insurers now have.
The Importance of Data-Driven Real Time Decisions
There a three fundamental questions that insurers should ask themselves when looking to write a risk.
Am I allowed to do business? Some of the insurers I have spoken to over the last few years are carrying out their compliance checks on the business in batch post sale. If you have the ability to understand who those UBO’s or Directors are in real time, you can carry out those checks in real time and get rid of those costly post sale processes that do nothing for the customer experience.
Do I want to do business with them? Risk selection is critical. If you have stakeholder related data and the information about the business you can make a better determination about propensity to claim.
At what price? With enhanced insight, your pricing will become more accurate
All this can be achieved in 5-7 seconds. Let me give you an example.
I was talking to an insurer a couple of years ago, and their biggest multi-million-pound claim came from a Phoenixed business, this is where business A folds and comes back as business B the next day. The directors are the same, the business activity and the location are the same. It is effectively the same business, but business A had lots of claims and business B can declare it has none. Wouldn’t it be good to be able to piece that together as you are making a decision? You can then make the right risk selection and pricing decisions.
The Fraud Prevention Use Case
It is fair to say that technology can help counter fraud activities across both underwriting and claims. We are not talking about just using the business and stakeholders as part of the network, it can be any subject or object of a policy or claim.
Talking about claims fraud first, it is well known that this type of network analytics can help spot organised fraud rings. But given the correct data, claims farming and other fraud methodologies can be spotted.
Slightly off topic, we see significant opportunities to counter ghost broking in real time with this capability. From a commercial underwriting perspective, having information on the involvement of directors in previous fraud cases in real time as you make a decision is a precious asset.
Beyond tackling fraudulent claims the opportunity to see the extent of a customer’s relationship with an insurer is an extremely beneficial insight. Imagine you have a small value business travel claim and the company has other policies, if there are shades of grey about the claim, you have more information to make a business decision.
Preventing Bad Outcomes For Brokers & Customers
When I was a broker, I saw a couple of instances that were, to put it bluntly, a bit shady. I’m going to try to keep this as anonymous as possible, but I promise these are real examples. There were two businesses I visited for the first time, the first was an industrial specialist cleaning company, and the second business made machinery that handled plastics. During my fact finding when I asked them where they operated they reeled off a list of places that covered just about every exclusion in a standard policy.
The cleaning company operated in mines, railways, and even culminated in cleaning an oil pipeline between the rig and the shore and the second company had a machine in the hot zone at Sellafield Nuclear power plant. Both businesses had a standard policy with all those exclusions from leading insurers. So they weren’t actually covered, and it is these sort of outcomes in the event of a claim that can give insurers a bad name.
I want to help insurers and brokers avoid these bad outcomes. As we look to bring our capability from the US to the UK this is the type of situation we want to help with. I see insurers using this type of technology to help the broker and the customer ensure that the insurance they buy is fit for purpose and covers them adequately. If we can encourage the free flow of data between insurer, broker and the customer surely that’s a good thing.
Three Learnings
Start with the value
It is change management
Have an incremental & agile approach
Start with the value. What are you trying to achieve, do you want to sell more faster or identify the risk of potential fraud in applications or claims. You need to tightly identify what your focus is and then communicate it effectively to avoid the “Computer says no!” mentality.
Align your goals with the strategic plan. Unless you have that alignment you are unlikely to have the support you need to follow through.
Adopt an incremental/agile approach. Identify the biggest problems and implement that first. Once you start to achieve a return and have proven methodology you can iterate and expand to other areas. Don’t try to do it all at once. For example, if you are not in control with the KYC – then start there, and later add underwriting insights. If you are losing business because you are too slow, automate the risk screening process. If high losses / potential fraud iare not being caught, setup fraud prevention at application or in claims.
About FRISS
FRISS is the leading provider of Trust Automation for P&C insurers. Real-time, data-driven scores and insights prevent fraud and give instant confidence and understanding of the inherent risks of all customers and interactions.
Based on next generation technology, the Trust Automation Platform allows you to confidently manage trust throughout the insurance value chain – from the first quote all the way through claims and investigations when needed.
Thanks to FRISS, trust is normalized throughout the organization, enabling consistent processes to flag high risks in real time.
Further reading:
Commercial Insurance – Fighting Fraud
With One Hand Tied Behind Our BacksCommercial Underwriting.
Do you really Know Your Customer?
