nicolas leroy

About Facebook, shopping engines and social interactions

April 22, 2010

This article is part of a series discussing what could be the perfect CSE for 2010. Following discussions on the core business of shopping engines (product search, product discovery, where to buy), I wanted to discuss the question of social interactions on CSEs… But after several weeks of thinking and numerous iterations of this article, I haven’t succeeded at organizing my ideas and defining what directions CSEs could take to build a successful social strategy… But, the announcements Facebook did yesterday during its f8 conference for developers have clarified some ideas…

About Facebook and the open graph

I’ve never been a big fan of Facebook (or similar social networks). While I understand how fun it could be to exchange with your friends using Facebook, it’s not in my habits to expose too much of my personal life. That’s why, while I’m quite engaged on Twitter or LinkedIn which I consider as professional networks, I’ve always been very careful with Facebook.

But let’s be frank: Facebook for now is the current winner among social networks (at least in western countries), any brand defining a social strategy has to consider Facebook (including CSEs), and the potential for CSEs is huge!

During the f8 conference, Facebook highlighted a very ambitious and disruptive product strategy. A quick summary of this strategy by SearchEngineLand (emphasis is mine):

[…] the vision articulated by Facebook CEO Mark Zuckerberg and Bret Taylor, formerly of Google and Friend Feed (acquired by Facebook), is of a more social internet, where relationships between people and things replace links between pages. The vision represents a shift from a Google-centric internet comprised of billions of unrelated documents and sites to a Facebook centric one where social relationships and affiliations are the connective tissue in a vast network.

The “Open Graph” — in which there’s information sharing between Facebook and other sites — reflects a hugely ambitious vision that began two years ago with Facebook apps and extended to Facebook Connect and now evolves to a larger initiative that makes Facebook the hub. A overly simplistic analogy would be: Facebook is the Sun and other sites are the planets.

[…] Publishers and developers can add a Like button to any page on any site. Adding that button effectively turns it into a Facebook fan page.

[…] By clicking “Like” (and being logged into Facebook) you transmit that you like a business, in this case, back to your profile and feed. But you might equally Like a band (on Pandora) or Like a news story (on CNN) or a movie (on IMDB). Those Likes become part of your identity and in turn part of the data available to other publishers and sites in the “Open Graph” that Facebook envisions.

[…] You must still authorize the sharing of your digital identity and public information with publishers. I asked Facebook about the following scenario: A user on Yelp “Likes” a sushi restaurant; can Urbanspoon later access my Yelp Likes (e.g., that sushi restaurant) to personalize Urbanspoon? The answer is yes.

From a walled garden three years ago, Facebook has now embraced a kind of openness: the web will now feed Facebook with activity data, data that will be redistributed to developers through APIs… Of course, this raises a lot of questions in terms of privacy, in terms of reliance on a single company; I’ve read that Facebook could become the single point of failure of the web; that Facebook is the new Google… But it is also exciting as it is a bold move, a way to make search (and Google) less important than today, a way for publishers to easily add a social layer on their sites…

Among the numerous articles covering the F8 conference, I have noted the situation experienced by Bret Taylor at FriendFeed and reported by TechCrunch:

[Bret Taylor] noted that at FriendFeed they found that the key to getting users to stick around and keep them using the site was that they had to connect with five friends. Unfortunately, when you’re a startup with not very many users, that’s extremely hard to do (yes, even just five). So FriendFeed implemented all types of logins and email contact lookups to try and help users find friends. The key to FriendFeed’s growth was Facebook Connect, as users were four times more likely to become engaged users if they signed up through that service, he said. In fact, if FriendFeed has continued on as an independent service, “we would have removed all those other signup buttons,” Taylor said. Yes, that includes Twitter and Google.

And also the launch of a new Credits program:

[Facebook] just announced a new Credits program called “App2User,” designed to enable merchants and loyalty program operators to convert their different types of rewards points into Facebook Credits.

Here’s how it would work: a rewards point system operator, like Chase, might offer their members a chance to convert their Chase points into Facebook Credits. The motivation for Facebook is to pump more Credits into the system (and users with Credits), so that they can spend more in apps. The motivation for potential partners would be to offer an attractive new option for users to spend their rewards points, in addition to existing options, like flights or flowers.

In other words, Facebook is creating yet another alternative payment option for Credits through this rewards points liquidity program.

Shopping engines and the social thing

CSEs hadn’t waited for the rise of Facebook not even for the emergence of the web 2.0 / social web concepts to implement social features: reading and writing product reviews was and is still a popular way to exchange between consumers… However, no CSE has really succeeded to reach the next level of user engagement despite numerous initiatives:

While the goal of CSEs is to redirect users to merchant sites as soon as possible, social destinations try to maximize the time spent on their sites… and I think this main difference due to distinct business models is the reason why CSEs haven’t succeeded to build successful social strategies.

Should social matter for shopping engines?

The challenge most CSEs have faced for the past three years is an eroding traffic, a difficult situation to build for the long term. There are several reasons that can explain such decline: Google changing the SEO rules year after year, users getting more mature and going directly to Google or retailers’ websites, no stickiness for users landing on CSEs….

I believe introducing social features has been seen as a way to stop this decline… A way to get users more engaged, a way for those users to bring user-generated content that could benefit the SEO strategy… Those are reasons for failure, as you first build social features to reach your business targets, not because you have a product strategy.

For me, social is not the silver bullet to have engaged users, but a way to improve awareness for shopping engines.

There are various kinds of shoppers – fashion addicts, researchers who take months to take a purchase decision, impulsive shoppers… But all those users at some point transform the shopping experience into a social activity: they discuss with their friends what products they would like, what products they find good or bad, they listen to friends who have advices…So it makes sense for CSEs to consider the social web, but Facebook is the platform for that, not the CSE!

CSEs should rely on Facebook to build their social strategy

Here are some ideas how I would implement social features from scratch on top of an existing CSE:

Summary / Final thoughts