The social network says acquiring Gnip will help companies better understand what consumers and other brands are saying across Twitter.
Performed properly, personalized product picks can power purchases.
Managing Editor, Mobile Commerce
Topics: Forrester Research, Forrester Research Inc., Patagonia, personalization, personalization technology, product recommendations, Recommendations, recommendations technology, Retail Systems Research, Retail Systems Research LLC
A shopper at Patagonia.com who's considering a pair of board shorts may or may not be in the market for other apparel or accessories to go with the shorts. Patagonia, however, doesn't take any chances. It prominently presents in a box other products, such as three trendy shirts, that its automated product recommendations system concludes may be a good cross-sell, based on what that shopper and other Patagonia shoppers have viewed and purchased.
The merchant started using the system two years ago and while it declines to reveal exact figures it says personalized product recommendations have more than paid for themselves. And they have boosted the conversion rate.
"Using product recommendations is what a direct marketer wants to do—eliminate the noise and speak directly to the customer," says Kevin Churchill, director of merchandising at Patagonia. "The online landscape has changed quite a bit, and if you are selling products on the web you need to have that edge and talk to your customer. Not only will it help your sales, it will provide better customer service and give them what they want and what they are interested in."
An increasing number of e-retailers are using personalization techniques and tools to create a more relevant shopping experience for consumers. 32.6% of e-retailers in the 2010 edition of the Internet Retailer Top 500 Guide used personalization on their e-commerce sites. That figure jumped to 50% in the 2011 edition of the Guide.
Personalization can help merchants better merchandise all the products they offer in a way that targets consumers based on a variety of factors, including, among other things, what an individual shopper previously purchased or what all of a retailer's shoppers viewed the most. 61% of retailers said personalization is among the most important merchandising tactics in web retailing, according to the 10th Annual E-tailing Group Merchant Survey, which canvassed 200 retailers of various sizes and selling in a wide array of product categories.
"Personalization is critical, essential, and growing in importance because as merchants really want to grow conversion, giving the customer a targeted experience through personalization is more effective," says Lauren Freedman, president of The E-tailing Group Inc. "It is a relevance issue—the more relevant the product, the higher the conversion."
And some experts say personalization is a must based in large part on how well online retailers know their customers.
"You have so much data on customers, it would be foolish not to take advantage of it—it's a lost opportunity if you don't take advantage of personalization," says Sucharita Mulpuru, vice president and principal analyst at Forrester Research Inc. "Product recommendations increase average order values, they increase time on site, they help expose consumers to products they might not have found otherwise. If you look at the most classic examples, Amazon.com and Netflix, it is about exposing a deep catalog of products and content that people might find useful."
And Forrester Research has found that the majority of consumers like it when an e-retailer recommends products. In a recent survey of 636 online adults who purchased a recommended product, 62% found product recommendations useful, 32% were neutral and only 6% gave recommendations a thumb's down.
Product recommendations today are the primary personalization tool. They help a retailer reach out to consumers through the pages of its e-commerce site to provide an experience that speaks to those consumers and draws them in. Merchants place them in key locations throughout a site: the product details page, the home page, a category page, the search results page, the shopping cart page and the order confirmation page. Some merchants place them in e-mail marketing pieces, and a handful even include them in online display ads.
Patagonia places recommendations on the product details, checkout and search results pages. Many executives in e-commerce associate recommendations with product and checkout pages, but not search results. Patagonia made the decision to intertwine search results and recommendations based on web analytics.
"We have a higher conversion rate when customers use site search," Churchill says, "so it made sense to put recommendations on a page where we have a lot of traffic."
Experts say placement is indeed key, and that there are best practices for personalizing an e-commerce site. Forrester Research suggests putting recommendations above the fold on a web site so consumers need not scroll down to see these effective sales tools. It also advises that the wording explain how product recommendations work.
"Instead of "We recommend,' try to expose the algorithm; for example, "Other people like you bought' or "Other people who have viewed this product purchased,'" Mulpuru says. "Show why you recommended it. If it's a highly ranked product or a lot of people have clicked on it, use that data to encourage people to click on it."
The magic number
Selecting the right number of products to recommend in a given space also is a best practice, says Nikki Baird, managing partner at Retail Systems Research LLC. But she says she knows of no "magic number," and that testing can determine what is best for a retailer.
But there are risks that come with personalization. If the focus of personalization becomes too sharp, a merchant may be missing out on other selling opportunities, experts say.
"Personalization is an understanding built over time," Baird says. "You have to be careful because just because somebody is exhibiting a certain behavior doesn't mean that is their preference all the time."
She suggests including a page as Amazon.com does where customers can refine their product preferences, such as by marking products that were gifts for others. That way the recommendation system does not take those products into account and keep presenting a man women's blouses because he once bought one for his wife. Including a checkbox near recommendations that reads, for example, "I like what is being recommended," can help refine results, she says.