It’s not possible to get a high level of accuracy for property insurance from the question ‘what’s the address?’ … Or is it?
A team that has brought a multitude of big data sources together in what it has called a “data supermarket” suggests that it is.
At its recent big data briefing, insurtech firm WhenFresh explained how its API uses more than 200 private and public data feeds, from sources such as Airbus, Zoopla and Bluesky among others, to autofill inbound insurance quote forms using just the address of any residential property in the UK.
Mark Cunningham, CEO and co-founder, said: “The idea of the API (application programming interface) is to composite data sets from a multiplicity of data sources and make it easy to use, so you could be looking at any aspect of form filling, or enrichment or looking at book value or reinsurance.”
Information can include the type of building, number of rooms and when a house was built, as well as the soil type it is built on, whether it is likely to flood, and the height and proximity of trees and other buildings nearby.
The company launched its big data insurtech direct to insurers in April 2019 and Cunningham said the tech firm had already done integrations “in less than half a day with proper blue blood insurers”, adding, “this is easy to use stuff”.
Speakers at the briefing, from a range of WhenFresh’s data providers, explained how different data types can spot and prevent fraud and offer a deeper understanding of the insurance risk in real time.
Richard Donnell, Zoopla research and insight director, said the property marketing, software and property attributes data business supplies its listings data to WhenFresh.
“It’s a rich data set of property attributes, flagging when properties have been listed for sale or rent, there’s a whole range of triggers and flags that you can do around that data.
“As I see it, the property listings is the tip of the iceberg and is part of a much bigger data set that we’re looking to explore as a business around leads and consumer interaction, understanding what’s driving demand for property, and the likelihood of it to sell.
“The big data scientists at Zoopla are now building up all sorts of AI and machine learning models around property descriptions to model forecasts.”
Donnell said Zoopla provides around five or six years of history and reasonably good coverage. But he added: “There’s a lot more we could do subject to whatever demand there is for more of this information and things that correlate to what drives loss around insurance and that sort of thing.
“We’ve got a pretty good track record of understanding of collateral value and home valuations with our Hometrack business, which is a data analytics business leveraging data around collateral values.”
CLS Data Limited provides data around transacting or holding property meaning they can map risks from the title of property and its physical characteristics.
Andy Lucas, marketing development director at CLS, said: “[The data] allows you to drill down, tracking through various titles linking into Companies House data to understand the directorship, and is it a phoenix company sitting behind this transaction?
So that automatically gives you some triggers for fraud detection, it allows portfolio analysis to link in the financials of the ownership to see if [the owner] Mrs Brown isn’t Mrs Brown at all and she is transacting on behalf of somebody else and is a potential fraudster.”
Also speaking at the briefing, Rob Carling, senior business development manager at BlueSky, highlighted how data from his firm’s aerial photography, LiDAR, thermal mapping, tree mapping, Ordnance Survey and GIS, could help insurers manage subsidence risks better.
He said trees are a big influence on house subsidence as they suck more moisture out of the ground in hot weather. Looking to the future he added: “We think aerial photography will go 3D. This means you can fly around the property and do fly throughs. This is the way this data is moving.”
Speakers from Airbus, Fusion Data Science, Cranfield University and Terrafirma all highlighted how the data their organisations provides to When Fresh helps to build a richer more accurate picture.
From Airbus’ big data sets from its satellites mapping flood risks, subsidence and soil types, Cranfield University’s in-depth study of the 1,000 different types of soil in the UK and what it means for the houses built on them to Terrafirma’s mining, ground and environmental information and Fusion Data Science’s software that models the world in 3D using AI to calibrate sensor data and satellite imagery, all the data providers help build a clearer picture of risk, one which looks like it is only going to keep getting more accurate.
Article reproduced courtesy of Intelligent Insurer. To login to Intelligent Insurer or to claim your 2 week free trial click here.
The Bank of England published its new Monetary Policy Report on 7th May 2020 – supported by WhenFresh data. WhenFresh have worked with the Bank of England (BoE) for many years our data and analytics have featured in BoE research & White Papers on several previous occasions. However, in the face of the lock-down due […]
Firstly, the WhenFresh team would like to wish that you and yours stay safe and well in these challenging times. While we are in uncharted waters, we are pleased to report that at WhenFresh, as a 100% digital data organisation, we are not far off ‘business as usual’. The nature of our business has meant […]
07 Feb 2020: Joanne Atkin of Mortgage Strategy Gazette reports on why M:Qube is different… M:Qube is about to launch into the mortgage market but it is no ordinary lender – it’s a fintech business first and a mortgage lender second… … Tony Connell, chief product officer, with a tech background at FICO, Visa Europe, Experian […]