Companies that apply a predictive machine learning model to advertise to customers could double their conversion rates, according to research from travel comparison website icelolly.com.
In the first research of its kind in the UK, icelolly.com worked in partnership with boutique agency Jellyfish to re-target more than five million buyers through two separate campaigns.
Launched over the course of a month, a predictive campaign used the latest in advertising technologies to classify users into buyer and non-buyer categories, forming a set of indicators to predict future behaviour. A controlled campaign was also used to provide a base set of results, engaging a random segment of users.
In total, the predictive campaign generated 146 conversions, more than double the 71 seen by the controlled campaign. Overall engagement from customers also increased, with four per cent more likely to click on adverts directly targeted at them.
The research has significant potential for companies when it comes to targeting future customers. Through rich data on consumer behaviour received from the predictive model, users can now be broken down into several sub-categories, from high-likely to buy to non-buyer, enabling companies to be more efficient and save resources.
Ross Matthews, chief marketing officer at icelolly.com, says, ‘These findings highlight the importance of providing a personalised service, with customers wanting to feel their needs are being considered. By using a predictive model, companies can better target and cater for their customers, which will have a significant impact on both feedback and revenue.’
Leeds-based icelolly.com has focused on its tailored offering in recent months, expanding into the curated space with the March launch of its ‘Top Deals’ product, a set of exclusive handpicked offers that provide extras such as room upgrades or guided tours.
The company, which was recently named in the Top 100 ecommerce companies in the North of England, handles over 130 million offers a day from some of the UK’s top travel companies. This covers beach, city, ski and cruise holidays, with more than 20 million visitors to its website each year.
Matthews adds, ‘Machine learning has the potential to unlock considerable growth if used effectively, and will prove crucial when looking to meaningfully engage customers and increase sales. We are committed to putting the customer at the heart of everything we do, and introducing the latest technology will be key as we make booking a holiday an enjoyable and hassle-free experience.’
Further reading on machine learning
