• by Steve Smith , Staff Writer, December 10, 2010

A weird YouTube phenomenon has cropped up this past year in the beauty and cosmetics category of uploaded video. Women started uploading videos of themselves showing off what they had just purchased at the mail. Dubbed "Haul Videos," the sub-genre usually shows someone pulling things out of a shopping bag and briefly discussing what they bought and why. Odd, yes, but the retail and cosmetics industry has already taken notice and Haul videos have now become commonplace. Well, it seems safer than skateboarding off garage roofs. But it also suggests a fundamental shopping behavior: seeing what people like you also bought. Why not use social networking to build that effect into the shopping experience?

We are still in the early stages of understanding how fundamentally the digital social graph can affect just about any mode of online targeting. The matrix of acquaintances that Facebook, LinkedIn and Twitter hath wrought can be tapped in many ways that are just waiting to be discovered. Brands have been trying to leverage the graph via their own Facebook pages and getting brand loyalists to share their preferences via wall postings and shared links. But what if you took your social graph with you into online stores to see which of your buds have also perused these shelves -0- and to check out what they purchased?

That is the model for TurnTo, a suite of "social commerce" tools that online stores can layer into their shopping experiences. If users opt into using Facebook Connect or other social network tools from LinkedIn, Plaxo, etc., when they start to shop, the consumer can see which of their friends have also shopped in the store -- and even sometimes what they have purchased. The system uses opt-in prompts at different stages to ensure privacy. At checkout, buyers are asked if they want to share their purchases with friends who may stop in - and according to TurnTo CEO George Eberstadt, up to 15% of these shoppers will agree to be a reference for their friends when they come to shop. The approach works especially well with enthusiast communities like art hobbyists. TurnTo just partnered with enthusiast publisher F+W to bring the system into its e-commerce site for painters.

The model is really a kind of personalized behavioral targeting. The shopper is looking for guidance from like-minded folks, and this approach is designed to be more intimate and perhaps relevant to them than a typical recommendation engine or bestsellers list. "Once it is enabled for someone shopping on the site, they can see who in their personal network has shopped from that store and what they bought," says Eberstadt. "If you see friends have bought from the store, you are more likely to shop there."

Interestingly, Eberstadt has seen the social commerce effect vary by site and category, and for different reasons. TurnTo has clients in the high end jewelry market that have seen significant life in conversions when this layers is employed. The vendors tell Eberstadt they think that the social graph helps validate the store. "Buying a $2,000 watch online requires confidence in the seller," he says. But even on the lower end, for a recycled toner cartridge site, they see high end performance. In this case, a category peppered with no-name brands, perhaps the graph also helps confirm that the site is not a rip-off.

For the F+W site, TurnTo claims that users employing the social commerce engine converted at a rate 400% higher than baseline in October and 300% higher in November. Across the company's partners, Eberstadt says that typical lifts are 2X to 3X. So if a standard conversion rate at an e-commerce site is 2% to 3%, layering in the social graph could lift it to 6% on those who opt into the system.

But of course the real issue becomes scale - and how much social networking overlap there is among shoppers at a given store. And then add to that challenge needing to get enough of those people to opt in to sharing. "The share of people opting in is not as high as we would like -- but higher than you would think," he says. TurnTo uses several methods of extending the network to achieve some scale, like using a shopper's Zip code to find people in their geographic area who also shop there. And ultimately it sometimes doesn't require a critical mass of socially connected users to have the social commerce lift. "We find that it doesn't take many friend connections for people to get excited about it, especially if it is a small store."  

TurnTo is working with 40 sites, including Dog.com, Compuplus.com and Teavana.