Roundtable: How can data enrich the motor claims journey from FNOL to payments?

Data is becoming more and more critical in the world of motor insurance. While data enrichment is used as a standard to assess risk and formulate prices during the underwriting process, its use from first notification of loss onwards in claims is more hit-and-miss.

What datasets are you using to deviate from manual validation, and what benefits have you seen from using them?

Simon Eells: We have a new claims system that is very data-orientated, so we are now using a lot of data.

In every client meeting, we discuss a whole set of data. The dashboards we put together give us the drivers’ ages and what sort of incidents they are having. We can drill down and provide the client with a better understanding of their drivers’ actions and where the pinch points are.

The aim from FNOL to closing claim payments is to make that easier, keeping the costs down and making it a smoother transition through the process."

Simon Eels

We are using this to speak to our insurers and brokers. The aim from FNOL to closing claim payments is to make that easier, keeping the costs down and making it a smoother transition through the process.

Dean Witherington: As you build out machine learning tools – the triage tools – you quickly create a lot of data. When using them to triage between total loss and a repair, you suddenly have masses of data – structured and unstructured. We have built a data lake to stick it all in, but I’m still not convinced we get the maximum value from it.

As you go through the claims process, this is the direction of travel, with more and more data from particular parts of the value chain. We have half a dozen machine learning tools at the minute, and they are trying to link the datasets together. But learning holistically from the customers across all these different data pieces is challenging.

Nick Kelsall: It is about trying to simplify the process. It comes from joining up our underwriting proposition right through to the claims proposition, having that holistic end-to-end view, and then aligning that data throughout the journey, which has been hard to do in the legacy world.

We have created this holistic end-to-end dataset, and our ethos is still allowing an omnichannel claims service but not asking the customer twice for the same bit of information. We should give this information at the point-of-quote or renewal rather than during a claim.

Lawrence Tanner: From an MGA perspective, we struggle with these elements because we have so many different products – many different types of policies – and we go through a third party administrator. So, all our claims are handled through them, which means they don’t always have access to our underwriting platforms.

Streamlining data from underwriting platforms onto a claims system, which has been built by another company, gives us hurdles that we have got to get over.

Rob Hammond: We very much have those challenges as a broker – probably tenfold given the number of agencies we have compared to an MGA. We have an outsourced provider, and data migration is probably a soft landing for the customer. Once they have passed their data protection questions, they will presume any outsourced provider to be the broker.

It is about getting enough information across to the TPA to ensure there is that soft landing and that they are acting on behalf of the broker seamlessly.

Do you feel pressure to move away from manual, and where is that pressure coming from to do this?

Witherington: Customers are driving everything and partly the regulator as well. The customer wants a decision; they want it now and don’t want to be told a couple of weeks down the track that it is a total loss.

Customers are driving everything and partly the regulator as well. The customer wants a decision; they want it now and don’t want to be told a couple of weeks down the track that it is a total loss.”

Dean Witherington

Hammond: Validation should be the easy part regarding vehicle driver details. You should be able to put them in a system, but from our point of view, we must then work on overlaying the claims footprint of the insurer.

Tanner: Depending on what you are dealing with, you need a hands-on approach to make decisions on specific claims, but your bread-and-butter claims can be automated faster than they are now. There are so many companies out there offering software and packages, but ultimately, it comes down to what the company saves versus the investment.

Eells: It is training the handlers about fraud, especially on the third-party side. We give them excellent training, and I’ve seen our fraud referrals to our carrier shoot through the roof. With every FNOL, they look at the speed, size of the vehicle, and what actually happened. We are seeing a massive increase in referrals now. That is literally from telling staff what they need to do.

Paul Llewellyn: The insurers that use data will be the ones that get the best results. Handlers can handle a claim, but what they need is to look at and learn from the patterns they have seen. It is up to us to use all those datasets to advise those handlers. Data can help enrich the decisions you make. Let’s give them the information they need to make the best decisions.

How can data be used to improve the customer journey?

Witherington: You can make a BACS transfer the same day, so the money is in the bank account shortly after the FNOL. That is a sweet spot if you can get to that position. And – with the volume of data we have about the customer, validation of circumstances and the quantum – that can start to come together well.

The more you get to understand the customer, the better. Whether it is social media or looking back through their history, you know what they want at the point-of-claim. Will they want a replacement? Fulfilment? Cash? Once we know this, things will become easier.

If you are talking about the customer journey, a customer doesn’t want to repeat their information for the replacement vehicle provision, which always happens. That is particularly frustrating for clients.”

Rob Hammond

Kajal Vakas: It is an education piece –because customers know you can get the money to them almost ASAP, as many insurers have the necessary technology.

Hammond: If you are talking about the customer journey, they don’t want to repeat their information for the replacement vehicle provision, which always happens. That is particularly frustrating for clients.

Llewellyn: It depends on the complexity of the claim as well. Once everything has been agreed upon, a straightforward claim should be quicker.

Kelsall: The only success measures that matter are whether we are controlling our loss and expense ratios and delighting customers.

And, if we do those things, we are in a good place as a business. The customers benefit from this as they are working with a profitable business that is growing; therefore, it can compete on price.

How close are you to getting that data to make the journey seamless? Is it a conversation you have with third parties about being more upfront with data?

Llewellyn: It is actively happening. We are talking about what we do with data, and giving the right people the access to that data, so all suppliers know what is happening. We send instructions digitally to our repair network, for example, so they can see what we can see.

It is vital to our strategy of having one single version of the truth, what is the best thing we can do for the customer, and it is more efficient, which makes us cheaper so we can pass that on to the customer. It is happening, and it is accelerating at pace.

Hammond: It is a massive challenge for us. We want to see more insurers feeding back into our systems. From our point of view, we feel it will benefit the insurers in terms of us not harassing their staff for their current position on this: the reserve, the payments, and the financials.

A few portals are knocking about and differ from good, to bad, to indifferent. The problem with most of them is the data sitting there, but you must go in and access it. You only know if something has changed once you go in and look for an update.

Eells: We have a portal with our carrier, but it all goes back to how the handler puts the data in there.

Tanner: Data quality is a massive problem. We have been doing quite a bit of work educating the front-end handler on the importance of what they put in the system at the front-end. It makes quite a significant impact on what comes out the back.

We need loads of data, so we create these extensive systems with lots of data fields that need populating. But we are reining in our data collection as we want less capture, more value, and more automation.”

Nick Kelsall

Kelsall: We need loads of data, so we created these extensive systems with lots of data fields that need populating. But we are reining in our data collection now as we want less capture, more value, and more automation.

We must ask ourselves: what is it being used for – and is it valuable? The last people who should be populating data fields are human beings or the claims handler. They should be running the claim.

Where do you think the motor industry is in terms of its use of data? How much further do you think it could go regarding data enrichment?

Eells: By bringing on the relevant TPAs, we are getting a lot more data, and we are using it a lot more, especially from our clients and brokers. But there is still a lot more that we could do as a company concerning data.

Tanner: As one of many start-up MGAs, we are a few years behind the big insurers. There are a few more hurdles that we need to get over. We also must factor in that we are starting from a dataset of nothing with no claims. We are building that portfolio. A lot more can be done, but it is moving in a positive direction.

Kelsall: It is about leveraging valuable data to drive things like ESG, with green parts. For example, how can we use data to find where green parts are, how do we acquire them, and then how do we get them to our network quickly so that customers are back in their vehicles quicker?

Hammond: We are moving into a phase of looking at where that service can be tailored to different tiers of clients and using data to look at that.

We are also measuring the data of other insurers and where they might surpass our performance. We need to offer our clients a consistent service with insurers across the board.

Witherington: We are good at the front-end of the process, but the Consumer Duty requirements will cause us to look further down that value chain, where we are weaker.

Moving that data downstream is becoming more important to us now.

Llewellyn: We can do so much more with data. But we have come on leaps and bounds, and I’m incredibly encouraged by the conversation around the table today. If we can get some consistency that can translate this information, there is a lot that we can do. 

Sponsored