Over the weekend, several of our fans alerted us to a blog post on You Exec. In it, You Exec uses FullStory to compare the quality of visits they get to their landing pages from different referral sources, and — guess what — not all sources are equal.

Alarmingly, they determined around 90% of the traffic they received from their Google and Facebook Ads campaigns were potentially robots, or at least low-quality users. Time spent on page for these shadow users was dismal, and conversions were an order of magnitude worse than You Exec’s other advertising channels.

We often do analyses like these ourselves and, judging by the 300+ comments on the blog post’s Hacker News thread, we thought more of you might be interested in doing the same. Here’s how to make it happen:

It all starts at the source.

We use query parameters to identify sources of paid traffic to various landing pages. To search for sessions where visitors arrive via Facebook, I just select Visited URL (query) > is > utm_source=facebook.

Then, like the folks at You Exec, I start watching sessions. If I’m interested in some quick analysis I can also take a look at the Searchies below the search results.

And I can add additional parameters to my search like Device > is > Mobile or Total Active Time > is at most > 5 seconds.

It doesn’t take long to get an idea of the kind of traffic heading our way from Facebook. 80% of it comes from a mobile device and the vast majority of these visitors spend less than 5 seconds on our site (yikes). You can take the same approach with Adwords, Twitter and any other places where you’re buying ads: just swap out “utm_source=facebook” with whatever channel you’re interested in.

So what gives?

Well it’s hard to say. We didn’t notice the same robotically erratic browsing behavior mentioned in the You Exec post. But, watching these sessions was still pretty insightful. We spent about 15 minutes watching sessions and our hypothesis is that most of our Facebook Mobile traffic has accidentally clicked on our ad while scrolling through their newsfeed. And those users probably hit the back button as quickly as their thumbs will let them.

After this realization, we went into Facebook to change our ad delivery from All Networks to Desktop Only. We’ll see how it pans out, but if the statistics we learned from FullStory hold true, we should see an uptick in Active Time — and, hopefully, conversion rate.

Of course, you might find something different for your ads and landing pages. And if you do, we’d love to hear about it!