Advertisement

Wine.com Boosts Order Value 15% With Custom Reviews, Recommendations

Capitalizing on impulse purchasing is not just relevant in the in-store environment, Wine.com employed customer reviews to leverage more than just recommendations. Recently named the number one online wine retailer by Internet Retailer Magazine, Wine.com uses reviews as relevant product information and content to support customer segmentation techniques.

Knowing that wine purchasing can sometimes be a “need to know but don’t want to ask” venture, the company is focused on providing lots of helpful information using what they know about customers, including browsing and purchase history.

Advertisement

The company wanted to enrich the experience of shopping for wine online and hone in on shopping purchase and patterns to “tap into the wisdom of the crowd.

“A good portion of wine buying is done for some-day consumption,” says Amy Kennedy, VP Marketing, Wine.com. “Expanding that to the online arena, we’re always trying to look at ways to enrich the shopping for wine online experience and add a lot more value than what consumers get at a typical wine shop.”

The company had its own homegrown solution that was limited so all visitors got the same recommendations without any personalization, because all recommendations were based on bestsellers. There was no definitive way to track reviews or their effectiveness.

Tapping into the Community
In May 2009, Wine.com implemented RichRecs from RichRelevance in under six weeks to deliver a more relevant customer experience by offering personalized product recommendations on product, category, home and search pages.

TheRichRecs solution is designed to leverage multiple recommendation approaches including collaborative filtering, personalization, analysis of current and past individual and “crowd” shopping behaviors, cross-placement optimization and a closed feedback loop to deliver the most relevant recommendations for each shopper at every moment. Automatic “competition” among 40+ recommendation types ensures accuracy and immediacy in the product recommendations.

Richrelevance software is dynamically learning based on people’s click and purchase behavior to find out what’s working for each particular customer. The solution offers users multiple capabilities, including an email module, retargeting module and specific recommendation module based on individual projects.

“The different modules were pretty turnkey to execute leveraging the upfront investment of time and effort to get the recommendations of the ground,” Kennedy says.

wine_com_cap_2

As a result, no two shoppers at Wine.com are recommended the same products, even when considering the same bottle of wine. Both recommendation types (such as “people who bought this also bought…”) and recommended products constantly evolve and adapt to provide a targeted and unique customer experience.

Recommendations are tailored to each shopper’s current browsing behavior and geographical location and incorporate the point of view of the Wine.com community – what they looked at, how they shop, what they buy – in order to expose shoppers to relevant products that they might not have considered otherwise.

With recommendations already driving a 15% increase in average order value on Wine.com—and as much as a 26% increase in recent weeks—Kennedy says the Wine.com team is pleased with the results of personalization.

“We offer a huge assortment of wines to an extremely diverse customer base of wine lovers who could be looking to restock their cellar, buy a statement gift, pick an inexpensive party wine, or just find something that they love to drink,” Kennedy says. “The idea of personalization has always been attractive to us but difficult to realize in our market. RichRelevance helps solve this problem and delivers an excellent customer experience.

Advertisement

Advertisement

Access The Media Kit

Interests:

Access Our Editorial Calendar




If you are downloading this on behalf of a client, please provide the company name and website information below: