Earlier this year PSFK spoke to Sonia Sahney Nagar, the CEO and founder of Pickie, about how brands are increasingly using social data to deliver personalized shopping experiences to customers. Pickie, an iPad app that curates a personalized shopping catalog for users based on their friend’s mentions of products on social networks, was tapping into Facebook’s ‘social graph’ potential long before Facebook announced ‘Graph Search.’
In an excerpt from our interview earlier this year, Nagar deftly answers the benefit of the new ‘Graph Search’ for brands:
You can build a very rich profile of users and their connections. For brands/companies/start-ups, at a basic level, this social information gives you the ability to quickly give users a great first-use experience. You can (and should) skip onerous onboarding processes to get users to relevant content more quickly. At a more advanced level, combining datasets you can start to discern things like affinity and ultimately influence.
While not the first brand to harness the power of social data, Pickie is a best-in-class example of how to properly leverage the information. Nagar and her team are mindful of the power- and limitations- of social data, careful to realize that access to social data is just one piece of the equation:
The data is what you make of it. Social data is good enough to provide recommendations with the caveat that you need to be intelligent about how you filter & surface the data.
There are a lot of product mentions that are ego-expressive vs. indicative of intent to purchase. Both serve a purpose in shopping. Ego-expressive, aspirational items often serve as the inspiration to start shopping. In this case, you may not want to buy the exact item featured, but might be inspired to shop similar items. Structured data lets retailers or social commerce sites create these product relationships (though not many do the work to do this today). In other categories, like books, the majority of sharing activity is around books people have actually purchased & read. In this case, for example, a ‘like’ of a book is powerful, and when presented properly, drives conversion.
Of note, I believe that social data is a starting point but it’s not a closed-loop system for making recommendations. There are shopping-intent indicators that aren’t captured at scale in any existing social network. We have our own smarts that let us sort, sift and prioritize that, and are proprietary to Pickie.
Nagar and her team approach social data in a smart and savvy way, using it to only direct, not mandate, their decisions. As brands not as familiar with social data as Pickie start to tap into the potential unlocked by ‘Graph Search,’ we asked Nagar her advice on how they can best approach the new search system:
There’s no doubt that Facebook has an interesting dataset, however I’ve posited before that data is only as good as the intelligence you apply to it. Creating a new search destination won’t be as simple as exposing Facebook data to free form search. I believe Facebook’s success will be determined by it’s ability to:
1. Focus on a specific search use case
2. Provide a comprehensive set of results for that use case
3. Add relevant data for the use case (mash-up their data with other data like location or product data, or create new data within Facebook)
The key value proposition of search is that it saves users time when there’s lots of information to digest. Thus comprehensiveness is important (otherwise you’ll need to search multiple sites to see all your options, which diminishes the time savings) and ranking quality is important (knowing that there’s some mechanism for putting the “best” results first will save you from having you to flip through all the results).
If I was Facebook [or a brand], I’d start with a free-form search box just to see what people were searching for. However, I’d quickly hone in on 1 case as a starting place, using the data to guide the decision on where to focus.
She also warns brands of the limitations of using Facebook (and social) data, especially in comparison to Google:
The problem Facebook faces is that it has a limited dataset; people only share a portion of their lives on Facebook. Therefore it is not comprehensive on any topic — even in knowing who matters to you socially, social connections are now notably fragmented across Facebook, Twitter and Pinterest.
Other search-based sites have competed with Google using the formula I outlined above (by focusing, being comprehensive within their niche, and adding relevant data to the experience). As examples, comparison shopping engines (Bizrate, NexTag, etc.) focus on products and and aggregate price and retailer information to provide value for users on a single dimension. Similarly, Yelp has carved out the restaurant niche by creating it’s own proprietary restaurant reviews and augmenting listings with useful restaurant information like location and contact info.
The juggernaut that is Google will continue to be a winner in the majority of search use cases because it more comprehensively indexes the world’s information, and has a beautiful virtuous cycle of searches feeding a ranking system to surface the “best” results first.
To further quote Nagar, while Facebook ‘has a massive base of users to gather data from,’ data from the social network will never be able to provide a fully comprehensive view of a consumer. Social data will be greatly beneficial in helping brands learn more about consumers, enabling them to better target consumers and provide them with personalized products and recommendations, but it isn’t a ‘magic bullet.’ Brands will need to use social data judiciously, and in tandem with other sources, to get the most informed view of consumers.
Thanks again Sonia!