How are insurers using data and technology to navigate the e-bike boom? 

The world’s Covid-19 pandemic experience of extended lockdowns drove an e-bike boom as people sought out new forms of exercise and transport. In the U.S. alone, e-bike sales doubled to USD 5.4 billion in 2020. In Europe, where many cities have better cycling infrastructure and more established cycling cultures, the popularity of e-bikes was growing steadily even pre-pandemic. In the Netherlands, for instance, one of the world’s most bike-friendly countries, e-bike sales have been increasing annually since the early 2000s. 

The global growth of e-bike users has also resulted in discussions over the sector’s growing pains: issues related to e-bike safety and collisions, fires caused by exploding battery packs, and policy debates over the rising speed of e-bikes.

Technology for specialised needs

Cyclists are traditionally protected by home insurance policies that protect personal property. Homeowners or renters insurance can cover theft and damage, first-party personal injury and third-party liability.

Yet the nature of e-bikes − which are more expensive, more high-tech, faster and heavier than traditional bicycles − means that conventional home insurance may not adequately cover theft or damage to e-bike parts such as batteries and electric motors. Similarly, they may not provide sufficient liability coverage in the event of accidents involving pedestrians or other vehicles. “Most policies exclude e-bikes outright – considering them motorised vehicles,” says Denis Voitenko, CEO of Velosurance.

Given the specialised needs of e-bike riders, who are a diverse cohort in themselves, insurers can harness technology, big data and analytics to provide better solutions and services to customers.

As Voitenko explains: “A baby boomer who just purchased their first e-bike to carry on the back of a camper van and a millennial who just gave up their car for an electric commuter bike will have different use cases, insurance experiences and customer service expectations.”

Gathering data

Effectively utilising and analysing a wide array of data sources can allow insurers to better serve both e-bike riders and the companies’ own goals.

Insurers can gather data from telemetry systems that automatically measure and transmit data via remote sources. Devices from e-bike sensors − which record speed, cadence and torque − GPS navigation systems and fitness applications, to those that track traffic and weather patterns, can acquire comprehensive information about policyholders and their usage patterns.

For example, many newer e-bike models come equipped with GPS, accelerometers, and 5G eSIM modules. Such devices can determine whether the bike has been in a crash, as well as key data such as the bike’s location, speed and direction of travel. “They can also determine whether an e-bike was left unattended for longer than usual or in a high-crime area,” Voitenko says. Some insurers use telemetry-based devices to observe users’ riding patterns. At the same time, wearable devices and fitness apps log real-time and objective data about a user’s exercise, health, fitness levels and activity patterns.

A scalable, retrainable system

This data can be fed into machine learning models that are highly adept at recognising patterns, detecting anomalies and forecasting trends to build predictive models. These can identify patterns that contribute to accidents or theft. Anomaly-detecting models, for instance, process data and identify unusual or abnormal data that could point to a potentially significant event, Voitenko explains. The models work at high-speeds and detect connections that could be easily missed by − or indeed imperceptible for − the human eye. And “unlike a claims manager or even a traditional analytical system, the model can be tuned, retrained and scaled on-demand,” he says. “When used correctly, these signals are tremendously valuable − from the cost-efficiency perspective and for customer nurturing, retention and overall experience,” he adds.

Such data and machine learning-based insights help insurers more accurately assess risk factors, adjust premium rates and pricing models, and develop personalised customer service, Voitenko says. As he goes on to explain: “An e-bike owner who commutes primarily on city streets at rush hour is at significantly higher risk of getting into an accident than a casual rider who uses an e-bike for exercise and stays exclusively on bike paths.”

Personalised service

Insurers can tap into technology service providers to optimise and streamline their data processes − or build their own in-house platforms. Machine learning-driven systems can continuously collect user data − from behavioural to demographic and more − enriching it and testing it to personalise customer service across all channels, Voitenko says.

By embracing technology, he argues, specialised insurance can ultimately “enhance risk assessment, minimise losses, personalise services, streamline claims processes, and proactively mitigate risks. These advances can lead to improved customer satisfaction, reduced costs and more effective insurance coverage.”