March 1, 2018 by David Gambrill, Editor-in-Chief
Using artificial intelligence (AI) to understand the likelihood of individual consumers making an auto insurance purchase, insurance broker Kanetix Ltd. was able to achieve a 13% lift on its call volume and lead generation.
“AI is something that the Insurance Industry has been talking about for a while,” Janine White, vice-president of marketplaces at Kanetix Ltd. says. “Being able to implement it not just for business growth, but also for enhanced customer experience, is very exciting.”
Kanetix Ltd. recently partnered with Integrate.ai to complete a pilot a project that uses machine learning to enhance customer experience and improve marketing ROI. Starting in September 2017, the pilot targeted consumers on the Kanetix.ca car insurance comparison website who were on the fence about purchasing auto insurance.
Based on the Kanetix website’s pool of data (including car information, the consumer’s postal code, accident history, current premiums paid, and even sales data), Integrate.ai configured the platform to predict which customers were likely to convert to a sale, as well as their preferred method of transacting, allowing Kanetix.ca to provide a tailored buying experience.
“When you go on our system and you get a quote, our AI system will score you,” Kanetix Ltd. president and CEO Andrew Lo explained to Canadian Underwriter Tuesday.
Then one of two things will happen.
“If you are a high-intent-to-purchase customer, we’ll just give you the quote and we know you are going to call us with a high degree of confidence,” Lo said. “We will eventually send you over to an insurance company, and the carrier will close it.”
But what about a score that suggests the customer is sitting on the fence? Maybe the AI system sees a data pattern that suggests the customer will not call, even when offered a deal.
“In that case, we provide a more personalized user experience for that person,” said Lo. “One experience is that if you call Kanetix, you get a $20 gift card. The AI determines that experience. It gives us a score and we determine what to do with that score. As a result of this whole system running, we were able to get a 13% lift in the number of calls and the number of leads to send to insurance companies.”
How does the AI system know?
“In its simplest form, AI looks at many signals,” Lo says. “One of the signals it looks at is price elasticity, so it will look at the final quote the person got. If they were able to save $500 or more, that’s a high-intent-to-purchase customer. That’s one example of a signal out of many.”
AI continuously learns from the data it accumulates, evaluating many data points to better predict a consumer’s intentions. “A programmer can’t write the software to evaluate those data points, it is that dynamic,” Lo said.
Over the course of the pilot period, the platform’s predictions improved by approximately 38 per cent. Kanetix.ca experienced a conversion increase in auto insurance transactions of 13 per cent, and improved its ROI by a multiple of 2.3 times. Kanetix.ca is now running 100 per cent of its Ontario auto insurance quotes through Integrate.ai’s platform.