Intema Solutions is set to launch its new digital catalogue technology in August that customizes users’ e-commerce experience based on their interests and behaviour.
The Montreal-based marketing software startup has made a business out of its predictive algorithms, which use advertisers’ and e-tailers’ first-party data to profile consumers and identify their most likely next steps on the path-to-purchase. So far it’s found success applying these algorithms to personalized e-mail campaigns for brands like Fido, National Bank, Desjardins and Aldo.
Digital flyers and catalogues offer a host of opportunities for personalization, which marketing automation companies like Adobe and Salesforce.com are developing technology to exploit. A case study performed by Facebook and Sport Chek in October 2012 showed that personalized “social flyers” doubled Sport Chek’s return on ad spend and increased in-store sales by 12% over the previous year.
To create personalized flyers, companies use algorithms to identify what items and offers each consumer would be most interested in and automatically redesign the landing page for each individual based on their unique preferences. The principle is the same for personalized e-mail flyers, advertiser homepages and online stores.
“In the end, you can send out a million e-mails, and get a million different pieces of content,” said Intema President Roger Plourde.
He said it was a natural step for Intema to apply its e-mail technology to digital catalogues that are hosted on companies sites, rather than distributed to consumers by e-mail. Much of the modeling and content design is the same on both platforms, except that catalogues rely on the consumer’s inbound interest, rather than outbound messaging.
What differentiates Intema’s solution from other personalized flyers and catalogues is its predictive marketing engine, Plourde said. The engine ingests first-party data from the advertiser about what a consumer has looked at and engaged with while they’re on the advertiser’s site or reading e-mail. The engine analyzes the data and predicts the next move that consumer will make – whether it’s a purchase, another visit to a company site or a newsletter subscription. Using these predictions, the engine builds a flyer that it thinks is most likely to drive that desired activity.
The kind of predictions the engine makes are based on statistical rules and machine learning, and can seem to contradict marketing logic, Plourde said. “Somebody could be buying protective felt pads to go under a table, and the system would predict the next purchase to be life insurance. There’s no real connection, but that’s exactly what predictive modeling does.”
But despite seeming counterintuitive to marketers, the results are hard to argue with. “When we send a campaign, normally we double results,” he said. “In one case, our client had between a 25-28% conversion rate – that’s sales – from one campaign. That’s an extreme, but we do achieve it.”
The company plans for the catalogue technology to be available by the end of the summer.
