Facebook tested a buy button last summer, too. Source: Facebook.

People don't like to shop on Facebook (META -10.56%). That's the conclusion to be drawn from Facebook's early tests with commerce.

But the company is back at it again, testing storefronts on select retailers' pages. Some of the product galleries include a "buy" button, which allows the entire shopping process to be completed within Facebook.

The new storefronts look similar to the product collections Twitter began testing last month. Twitter also enables the entire shopping process to be carried out without leaving its app or website. Pinterest similarly rolled out buyable pins, and Google istesting a "buy" button, as well.

But Facebook has one big advantage over all of them: its social graph.

It knows what brands you love
Facebook has done an excellent job encouraging users to follow brands both on Facebook and Instagram. In fact, Instagram may provide Facebook a better idea of what its users are interested in. As Facebook marketing exec Carolyn Everson has noted, Instagram users follow their passions; Facebook users follow their friends.

Interestingly, Twitter should theoretically have a similar interest graph as Instagram, but a failure on management's part to instill that behavior in users has left its graph with lots of holes in it. That makes positioning buyable tweets significantly more difficult. Likewise, Google might know you're looking for an umbrella, but it doesn't know if you'd be much more likely to buy a specific designer's umbrella if it showed it to you first.

It knows what your friends are buying
Facebook has seen tremendous growth in its video product during the last nine months due to the viral nature of its network. If a user likes or shares a video, a bunch of his friends will see it. And if they like it, a bunch of their friends will see it. And so on.

Because Facebook has complete control over its users' News Feeds, it can pretty much stick whatever it wants in there. So if someone buys a new dress through one of Facebook's storefronts, Facebook can show that to all of her friends. And those friends can like and share, or even buy, that same dress -- (although they may be better off browsing the storefront for a different dress).

This viral mechanism is something no other business can replicate. Twitter's reverse chronological timeline prevents it from accomplishing the same effect. Google's main service is a solitary experience. Only Facebook is in a position to make things go viral.

The biggest challenge to overcome
As Facebook's early experiments with commerce have shown, people don't come to Facebook to shop. Getting users to overcome the idea will be difficult, but it's not entirely impossible.

Facebook has already set the stage for more commerce by enabling Messenger users to send money to one another by linking a debit card to their accounts. It's also experimenting with businesses on Messenger to provide customer service and track orders.

Additionally, it's reportedly working on a concierge service -- codenamed Moneypenny -- that integrates with Messenger. All of these efforts are designed to get users used to interacting with businesses and spending money on a Facebook platform.

Getting retailers on board won't be a problem. Facebook has the largest audience in the world, and it dominates time spent on mobile devices -- where retailers are struggling to convert sales.

If Facebook can leverage its social graph to convert browsers into buyers, it could grab additional ad revenue. Facebook isn't charging a commission on sales, but brands looking to promote themselves or a new item in the store are more prone to advertise when they know Facebook can convert on sales.