Facebook's effort to tap into the popularity of buy, swap and sell groups, Facebook Marketplace, is two years old this week, and to mark the occasion, The Social Network has outlined some interesting new, AI-fueled tools to enhance the Marketplace experience, and boost Facebook's evolving eCommerce efforts.
First off, Facebook's providing sellers with more context as to what they should add to their listings to maximize results, based on image recognition and comparative insight from similar listed items.
"What does this mean exactly? Here’s an example: if you wanted to sell your home office chair, Marketplace could use AI to help you sell it even faster by suggesting you price it between $50-75 based on what similar chairs recently sold for. It will also automatically categorize the chair as “furniture” based on the photo and description, so that you don’t have to."
That might not seem overly ground-breaking, but there's a little more to it than this basic functionality.
For example, right now, Marketplace uses machine learning tools to highlight similar items to those you've searched for and/or expressed interest in. Up until now, those matches have been fairly limited, because they were largely based on simple text descriptions - and not everyone's great at explaining what it is they're looking to sell.
As explained by Facebook:
"From the very first search, results are recommended by a content retrieval system coupled with a detailed index that is compiled for every product. Since the only text in many of these listings is a price and a description that can be as brief as “dresser” or “baby stuff,” it was important to build in context from the photos as well as the text for these listings. In practice, we’ve seen a nearly 100 percent increase in consumer engagement with the listings since we rolled out this product indexing system."
So it's not just about ensuring that your listings are achieving best results, it's also building a contextual understanding model to better map the items on Marketplace, and then using that to link relevant items to other searchers based on more data points, which is far more significant in the scope the broader process.
Facebook's also adding some other, more subtle, AI improvements.
"One element now automatically enhances the lighting for photos as the seller uploads them, making the images easier for buyers to see what’s pictured."
An interesting, yet likely overlooked process, while Facebook's also (as you can see in the above video) providing new M Suggestions message prompts to streamline communication between buyers and sellers.
"For example, in a 1:1 conversation, when a seller is asked a question such as “Is the bicycle still available?” M can suggest replies like “Yes,” “No,” or “I think so.” If the answer is no, the product is no longer available, then M may also prompt the buyer with a suggestion to “Find more like this on Marketplace” with a link to a page of items similar to the one the buyer was asking about."
These are all small updates in a relative sense, but the performance impact could be significant. For example, Facebook notes that between 7% and 9% of sellers abandoned the listing process before the addition of auto-suggestions for price and category. Given the millions of items now listed on Marketplace, that's a significant chunk they can reduce with such tools.
But the real kicker of these new tools may be what's coming - as Facebook builds a contextual engine for products, which uses image identification to help classify each, they also, by extension, are building a visual search tool, similar to Pinterest's Lens and Snapchat's new camera shopping option, which searches Amazon's product catalog.
That's the next level for Marketplace shopping:
"We’re also testing camera features that could use AI to recommend products you might be interested in. Say you liked your friend’s headphones and wanted your own; you could snap a photo of the headphones and Marketplace’s AI technology could recommend similar listings for sale nearby. In future, AI could help simplify tasks like completing an outfit or home design project. For example, you could upload a photo of your living room and get suggestions on furniture to buy based on your layout and size."
Given the aforementioned Snapchat/Amazon initiative, which was launched last month, I do think the use of the term 'snap' in this paragraph was strategic.
Each platform's taking its own approach to similar eCommerce options - Pinterest's using its database of 175 billion Pins, Snapchat's tapping into Amazon's catalog of 560 million or so products. Facebook has millions of products listed on Marketplace, which would be the smaller of the three databases, and could limit its subsequent matching capacity. But then again, Facebook has also been slowly evolving Marketplace beyond just used items listed by users.
Indeed, Facebook has also added a dedicated section for vehicles which includes listings from traders, a section for home rentals, eBay 'Daily Deals' listings. While Facebook hasn't made big announcements about these additions, it is building out Marketplace into a more all-encompassing online shopping option. And with more than a third of people on Facebook in the US using Marketplace every month (over 80 million people) the opportunity is significant.
For comparison, Pinterest has 250 million users total, with only a small percentage of them actively using Lens. Snapchat has only just launched its visual search tool, but at 188 million users, it's also not likely to offer the same potential scale as Facebook Marketplace in this regard.
It doesn't seem like it, but Marketplace may be in the best position to capitalize on social eCommerce, based on usage trend data.
In addition to these new, functional additions, Facebook's also updated its rules to help ensure buyer safety through Marketplace.
The new updates include:
- Detecting and Removing Inappropriate Content - Facebook's now using new AI tools to detect and remove items which violate its policies "by analyzing the images, content, and context within a listing".
- Buyer and Seller Ratings - Facebook's added a new rating system for buyers and sellers to provide new ways for users to get quick context on what to expect (similar to eBay's ratings system).
- More Robust Reporting Tools - Facebook's also added more tools to report items which violate the platform's Commerce Policies. "You can also report buyers and sellers if you see activity that you think shouldn’t be on Facebook".
There are also new 'badges' for buyers and sellers who've confirmed their identity, along with features that'll make it easier for sellers to select a nearby public place to meet in person.
In some ways, Marketplace may be the sleeping giant of Facebook. While all the focus is on Stories and Messenger bots and other potentially immersive, money-making tools, the growth of Marketplace could offer even more options in this regard. And if Facebook were to add in, say, an option to search by user-listed items or Page listed, that could make it a significant brand consideration, especially as its visual search tools evolve.
There's money in eCommerce, and there's huge capacity for social networks to connect people with products. And Facebook, with the most users and reach of any platform, could win out in this race, especially if it can get such tools right.
Add to this its advancing translation tools, and there's a lot to consider in Facebook's Marketplace development.