All year I've been salivating over Facebook Instant articles, which provide a greatly improved mobile experience when clicking from the social media behemoth's newsfeed.
When implemented properly, a Facebook Instant article will load in less than 1 second -- compared to ~10 seconds (or more) for a standard mobile page load. That saves a user's time and significantly lowers the barrier to interaction with your content.
Of course, for many Facebook users, Instant articles were a subtle addition; they simply got a great mobile experience when clicking over from Facebook, but didn't know what it was.
When Instant first launched, Facebook only allowed certain companies to utilize it. (Most were big publishers like The New York Times and Washington Post).
In April, Facebook opened Instant articles to all publishers willing to do a little bit of leg work, and I rushed to be among the first to take advantage.
Does Facebook Instant Move The Dial?
Though the big publishers had been using Facebook Instant for almost a year, they didn't share any of their data with other content creators (no surprise there). Which left open the question: How well does it work, and what are the pros and cons?
I decided to test their impact by instituting them on BarBend.com, a project I helped launch that's quickly become a leading resource for strength sports news and entertainment.
To test Instant's impact, I alternated each published article on BarBend between Facebook Instant and traditional mobile experience so that I could gauge the impact.
To publish a Facebook Instant article, you need to first set up your site in a specific way and then connect it to the development tools on Facebook. It sounds intimidating, but Facebook has an open-source project to help WordPress users to easily publish from their blog.
There are still a few tweaks to work out with the out-of-the-box plugin, but generally it's pretty easy to get Facebook Instant up and running; I was able to have most of the kinks ironed out within 48 hours.
The Positive Impact (Numbers Included)
After three weeks of use, it was clear that Facebook Instant was outperforming the standard mobile experience by a wide margin. Overall, the posts resulted in 25 percent more shares and 32 percent more likes compared to traditional loading pages.
In its first full month, BarBend attracted just over 100k visitors, with roughly 80 percent of the traffic coming from viral Facebook traffic. Given that it was a brand new site without any PR spend or official launch announcement, our team was thrilled.
The bigger exposure also led to a fast growing Facebook page audience. Though I can't directly attribute growth numbers to Facebook Instant, our "likes" grew at a much faster rate than any page I had previously managed.
Becoming Facebook Dependent
Though Facebook Instant article stats look good, there are some major concerns going forward. The big one is that these articles are technically hosted on Facebook, not your website, so you have significantly less control over the experience; if Facebook wants to change something with the interface, they really don't have to answer to you.
While the mobile experience is great, you won't be able to do some basic optimization techniques you would otherwise use your site, like a popup to capture email addresses or a mobile bar so users can whisk through more content.
While the posts performed much better (and anecdotally grew BarBend's Facebook audience much faster), Instant articles also make a site much more dependent on Facebook traffic. In the short term, it's not a big deal, but as your site grows, a Facebook algorithm change could be devastating.
Though I do have some long-term concerns with Facebook Instant articles, when launching a new site, they're really a no brainer. You get significantly higher engagement and the opportunity to build your brand more quickly.
I've since implemented Google AMP and Apple News on BarBend, which are Google's and Apple's similar mobile-optimized experiences. In the first month, these didn't have any sizeable impact on traffic, but I think long term they could also have a big impact. I'll make sure to report back on that in due time.