Hello PAYG Part Two: AI, Ridesharing & Bias Free Data

OK, so here is part two of our Hello PAYG themed feature for March.

Let’s explore market opportunities, and challenges, beyond the PAYG or usage-based model replacing TPF&T or Fully Comp over time. Consumer choice is going to demand insurers are more flexible, but how can data be used without income, gender, postcode or ethnicity bias for example, or match the precise demands of ridesharing login/log-out times when it comes to policy admin and compliance?

Plus, there is a swarm of battery powered vehicles, of all shapes and sizes, which are going to enter the road network in the next 5-10 years. They will all need cover, often on a per-trip basis.

Rich Tomlinson from Percayso-Inform offers this view;

“True flexibility means that the profile of those wanting or who could benefit from PAYG insurance is many and varied, and their needs and details may change more rapidly over time than the average policy holder – for example, as a delivery driver borrows a friend or colleague’s car or car sharing amongst millennials becomes more prevalent. Times of use, purpose, vehicle as well as personal circumstances can all change – potentially on a daily basis.

But while this could be seen as a huge opportunity for insurers to meet a latent demand, it’s also a potential minefield when it comes to trying to accurately assess risk. This is where having access to dynamic tech and up-to-date and accurate data in real time matters, and it is vital that this should then translate into a frictionless, quick and easy process for the policyholder to change, update, switch on and switch off their cover.

The team at Percayso have been helping a number of insurers who are very active in the PAYG market. In our experience the successful ones invariably have a modern, dynamic, API connected, app enabled tech stack that makes things really easy for the customer and access to accurate, relevant data that informs risk and price in real time. Equally important, they have a forward thinking and customer centric mindset – right across the business, a real desire to disrupt and a confidence to know that customers will benefit from and increasingly demand the service.”


That is a key point in this journey; it isn’t enough simply to want to disrupt the market, whatever that means in different territories. You have to deliver benefits to consumers too. Then there’s the matter of product distribution which means that the disruptors often need the stick-in-the-mud giants to provide an economic route to market.

What that means in practice is that big names insurers and brokers in the motor, van or motorcycle market can partner up with companies who are already offering off-the-shelf solutions to PAYG problems. Projects like Lloyd’s Lab, or Zurich’s Innovation Foundry act as test beds for all sorts of insurance ideas, and associations like InsTech London can open the right doors for insurtech start-ups too. Fact is, everyone in this industry is looking for ever-quicker routes to market – and fail fast vs proof-of-concept is at the core of many a product development journey. Nobody wants to waste time or money, despite the billions pouring into insurtech brands via funding rounds.

Andre Symes from Genasys Technologies comments;

“It takes significant investment – both time and money – to adapt existing platforms whether that’s for a major existing provider to launch their own PAYG offering, or to accommodate the tech of a new insurtech partner who has the tech to get the product out there. The most cost effective way forward is to buy an off-the-shelf software as a service solution to launch the product, properly test it and see if it will achieve the desired scale of take-up. If it does, then any organisation can justify the cost of either retaining the bought-in system and running it alongside their own core platform, or reprogramming their own platform to cope.”


Here is another interesting trend; public sector ridesharing projects. As big cities around the world are following the World Economic Forum mantra of `Build Back Better’ by banning private car use, or charging ULEZ tolls, that opens the door for ridesharing on-demand, often via e-scooter or e-bicycle as an urban mobility option. Then there’s the whole Covid Project Fear thing, which will probably keep many people well away from buses, trams or trains for a few years. It’s an opportunity for car insurance brands, but comes with challenges, not least user ability, helmets, licences etc.

Tomer Kashi, CEO of Voom offers this take;

“Micro-mobility could surpass public transportation within cities in the coming decade. Ridesharing is already an accepted and normalized way to get to and from work, and a majority of total passenger travel in China, Europe and the US is in trips shorter than 8 kilometers (5 miles). Workplaces have started to designate parking spots for e-scooters and bikes, especially in warm, urban environments.

E-scooters have caught on to the point where common caricatures about who rides them (young people otherwise attracted to new mobility) have faded, and governments are noticing. Last month, two US congressmen introduced the E-BIKE ACT, which aims to offer $1500 in tax credit for e-bike purchases. Owned vehicles will not disappear outright, but will lessen and shift more towards leisure. 

Because we are beginning to use our cars for different purposes, our relationships to these vehicles will invariably change. This includes how riders and drivers maintain their vehicles, how dealerships and marketers sell them, and how insurers harness new data sets to offer more flexible options for coverage.”

Your insurance brand is being weighed and judged every day, by social media activists who love to find grievance. So build your AI systems free from unconscious bias.


The thing about using stacks of data to calculate risk is that you could pick and choose how much weight to attach to a particular layer of the data sandwich. In a woke era, where Prince Haz can blag a job as a fake news expert, even though he’s married to a fake news generator, you cannot underestimate the power of the online activists. So transparency is key when it comes to using AI to underpin every aspect of your product; risk, quote info formats, pricing, claims criteria – the lot.

Dr Nicolai Baldin, CEO and founder of Synthesized notes,

“The reputational risk of all organisations is under threat due to biased data and we’ve seen this will no longer be tolerated at any level. It’s a  priority now and must be dealt with, both from a legal and ethical standpoint. Synthesized’s Community Edition for Bias Mitigation is one of the first offerings specifically created to understand, investigate, and root out bias in data. We designed the platform to be very accessible, easy-to-use and highly scalable, as insurance organisations have data stored across a huge range of databases and data silos.”

In the future insurers and brokers need to demonstrate that ther datasets don’t use outdated models like postcode discrimination, income, marital status or job based pricing. Such 1990s call centre box-ticking is definitely going to look like unconscious bias if you simply transfer those old fashioned prejudices to your shiny new PAYG policies. Cover has to be based on real-time facts, not historical assumptions.


Crispin Moger, CEO at Marmalade has seen how younger drivers modify their behaviour to win a decent score, calculated by a simple telematics windscreen tag gadget;

“The policy works in tandem with our cutting-edge App & Tag technology. It can be managed through a smartphone and the telematics device simply sticks to the inside of the windscreen – which makes it simple to set up without compatibility issues. Drivers can check miles used and their ‘driving score’ on different journeys, which measures acceleration, braking, cornering, speed and phone use. These events are marked on the journey route where they happened, so they can see where they’ve driven well and areas for improvement.”

Those young drivers can build up a scorecard over say 5-6 years, which offers insurers much more insights into pricing the risk long term. But ThingCo are expanding on the in-car telematics idea with Little Theo, which mashes up the dashcam idea, with a bit of emergency rescue and a layer of driver coaching on top.

Theo – a handy gadget to have on your side if the worst happens

Having a time-stamped data record of speed, braking, location, plus the voice activated conversation post-incident is something the industry can use to streamline claims payouts as never before. In essence, AI can automate many of the lower cost `fender-bender’ claims this way and although that doesn’t sound too exciting, it could mean big insurers save thousands every month on hundreds of claims.

By offering consumers lower premiums, plus rewards sweeteners and the capability of switching business, or temporary driver cover on and off, insurers can build truly flexible products that fit people’s lifestyles. The market will shift gradually from its traditional model of annual car insurance and the process won’t be easy for many brands. There is no doubt that some players here and now will exit this market in a decade, as private car ownership itself diminishes in the UK.

But the opportunities are also there for battery-driven mobility devices, many of which are being backed with grants from the public sector – they want this brave new transport world to succeed, with an almost religious fervour. That means money has to be spent to win those voters over, to compel them to switch to e-scooters, bikes and ridesharing battery powered city cars. That’s how politics works.

The green gravy train is leaving the (charging) station soon, so jump on board.







About alastair walker 6139 Articles
20 years experience as a journalist and magazine editor. I'm your contact for press releases, events, news and commercial opportunities at Insurance-Edge.Net

Be the first to comment

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.