J. Crew experienced a 5% lift in demand on its Web site when it improved its analytics. The resulting revenue was significant enough to cover the costs of starting up the analytics program, eight months ago.
“A year ago, we had a lot of information, but we needed actionable information,” said Jayson Kim, senior director of Web marketing applications at J. Crew, at an e-commerce analytics session at net.marketing conference in New York.
The analytics included a product-recommendation engine. When a customer was looking at pants, for example, the search engine would make suggestions about types and styles of slacks, then display compatable shirts. Finally, at check out, it showed an impulse item, such as a belt.
Tests showed that 8% clicked of visitors clicked on the recommended product, 6% added it to their shopping cart. Some 2% of customers added a recommended product to their cart.
Using cookie technology, buyer behavior was tracked. When first-time visitors were offered a promotion for free shipping, J. Crew got a 2% lift in overall business. Previous shoppers received an offer based on past business or geographic location.
Offers included discounts and incentives. Since the channel the customer shopped at was tracked, store shoppers would receive a coupon for use in a retail outlet.
Five segments were defined to target offers, including purchase history, frequency, recency, channel and preferences.
digiMine Inc. provided the analytics technology.
Net.marketing runs through tomorrow at the New York Hilton.
Network
