At the beginning of 2010, Facebook rolled out a number of significant changes to their platform, including most notably the introduction of the Open Graph protocol and Facebook social plugins.
Since then, retailers have been trying to determine the best strategies for taking advantage of these options. Two significant strategies (at least) have emerged so far, represented by the Levi’s Friends Store and the Amazon Recommendations page. In essence, Levi’s lets you like products from their brand, and shows you which ones your friends have liked, while Amazon instead reads from the Facebook profiles in your social graph and uses that information to drive specific recommendations.
Which is more successful? As always, it depends on what assets you have to leverage and what you hope to accomplish.
Do these jeans make my ___ look ___?
The Levi’s Friends Store, in essence, allows you to “like” any of the brand’s products, and if you are logged in to Facebook when you visit (or login during your session), can use your friends’ likes as a filter on what is shown in the store view. (If you’re not logged in, they can still show you what “everyone” liked on Facebook).
It’s easy to miss the significance here. Levi’s is letting me use my list of friends (which I created not on Levi’s store but on a third party social networking site) and what that list of friends likes (the record of which is stored at Facebook not at Levi’s) to filter the view of products in their store.
On the one hand, this might be seen by most brand managers as quite risky. What if my friends don’t like anything they offer? Won’t the things people like most get sold out, leaving only an island of misfit toys left in the nearly empty storefront?
On the other hand, given the massive popularity, longevity, and consistency of Levi’s brand, it also strikes me as quite safe. Who doesn’t find at least one pair of jeans at Levi’s that they like? Given that there’s no (official) “dislike” button, and there’s the fallback to “Everyone,” what’s really the danger?
More interesting to me is that liking a product in the Levi’s store has two other effects that I don’t think most people realize: it impacts your Facebook Profile, and it gives Levi’s an opportunity to write on your wall. Or, to put it in Facebook’s terms:
This means when a user clicks a Like button on your page, a connection is made between your page and the user. Your page will appear in the “Likes and Interests” section of the user’s profile, and you have the ability to publish updates to the user. Your page will show up in same places that Facebook pages show up around the site (e.g. search), and you can target ads to people who like your content.
If you’ve liked a product on the Levi’s store, check out your Facebook profile: on the info tab, in the “Likes and Interests” section, under “Show other Pages.” (You may even need to click on the “and more” link at the end of the “Other” section to see them all).
In fact there’s what seems to me a bug in Levi’s implementation. When you click “like” on a product, you may see a little error message like this one:
If you click on the red text “Error” you’re told in a Facebook style modal dialogue box:
This happens because the meta tags Levi’s passes to Facebook when you click the link button define the “type” of the object being liked as a “product_service” but that isn’t one of the object types Facebook supports.
Ultimately, this doesn’t really matter as Levi’s isn’t (I assume) as interested in getting their products listed in your profile’s “Likes and Interests: Other” section so much as they are in being able to display your face next to the product to your friends (and perhaps the ability to write on your wall and target you with ads inside Facebook, though I’ve not seen that yet).
Amazon knows what you like . . .
Reading the items in your profile, however, is exactly the part of the Facebook Open Graph protocol that Amazon relies on for their recommendations page. Rather than asking you to “like” products in their store, Amazon reads your profile and your social graph to:
- Create Recommendations for you personally, based on the books, music, and movies in your Facebook profile
- Inform you of upcoming birthdays by people in your social graph, and
- Make recommendations for those folks, based on the books, music, and movies in their Facebook profiles
Where Levi’s store reads people’s likes looking purely for items they’ve liked on the Levi’s store, Amazon finds things you’ve liked independently of where you liked them. To see yours, log in to Amazon and click on the “Recommendations” link in the top header – you should see a Facebook box in the top of the rightmost column – once you connect with Facebook it will begin to show recommendations customized based on your profile.
Amazon can show you what’s popular amongst your friends on Facebook — regardless of what site they used to “like” those things — as well as make recommendations for specific friends with upcoming birthdays, provided they’ve connected to Amazon as well (or set their security settings to allow that profile information to be public).
They can also leverage their existing recommendations engine, knowing for example that if I like Alison Krauss as a musician, I might like her CDs, or that people who liked certain authors might like certain other authors. Much of what Amazon is relying on here are the things users added to their profiles at sign-up — back when you could add books, movies, and music to your Facebook profile without first finding a page with a like button on it — and extensions based on its own knowledge of music, books, movies, and the relationships between people who like certain styles, genres, or authors/musicians/directors.
It isn’t just that they can tell me what is specifically in my friends’ profile, but that they can tell me what isn’t in their profile but probably should be, based on the items that are. They’ll even share with you why (on the basis of which other items in the profile) a given recommendation was made.
Levi’s can show you friends with upcoming birthdays, but the likelihood that they will find not only a friend who has used the Levi’s online store (and taken the time to like a product) but also one with an upcoming birthday are much lower than the odds that Amazon takes, since most Facebook users have at one time or another liked a book or a movie.
At the end of the day it’s a strategy well suited to a retailer with both a deep understanding of buying patterns in, and a strong presence in the market for books, movies, and music, which are the most structured parts of the Open Graph protocol and the most complete sections of many users’ Facebook profiles. (Unlike Levi’s “products and services” tag, which relegates their likes to the “Other” section of the profile, the books, music, and movies someone likes get their own specific sections).
What’s a retailer to do?
If you’re a retailer thinking about how to leverage the open graph protocol for your users, you need to consider what assets you bring to the table, what business goals you hope to accomplish, and what your users will be perceive as useful.
