Should publishers like Facebook’s new Like button? I recently wrote
about some of things publishers might not like about it.

One commenter who is a fan of the Like button took issue with them.
He asked, “…allow me to ask you whether YOU have quantitatively
measured the impact of the five risks that you are warning against?

It’s a fair question worth responding to, and while it is, of course, impossible to quantitatively measure, for every single publisher on a global basis, the impact of all of the potential drawbacks to the Like button, I happened to find an interesting post noting that some have looked quantitatively at one of the negative impacts of the Like button: page performance.

A gentleman by the name of Alex Sexton found a great example of a company using the Like button extensively on its website: popular jeans company Levi’s. The results:

From what I can tell, a single Facebook like button requires at least 12 unique requests and ~1 second of load time.

The button adds ~161 new http requests to this page on Levis which is ~4sec

Compare the difference in when onload kicks in via iframe compared to via JS.

Using Google’s Firebug Page Speed add-on, the impact of the 20+ Like buttons on this page is easily confirmed. In my own tests, a typical hard refresh of the page results in requests to the Facebook servers that continue for well over 5 seconds on average. Qualitatively, some of the Like buttons at the bottom of the page only starting to appear after the rest of the page has been loaded — not the smoothest experience to say the least. Note that this is all on top of pages that are already quite heavy with JavaScript, HTML debugging comments that should really be hidden from non-developer requests, etc.

That’s bad news because for a retailer like Levi’s, the potential financial impact of just an extra second of load time could be significant. According to a study conducted last year by Forrester Research, 79% of online shoppers surveyed said they would be less likely to make another purchase from a slow site, and nearly half of them expected a perceived page load time of 2 seconds or less. And performance expectations are increasing, not decreasing, which means that online retailers in particular need to be careful about what they add to their pages.

In the case of Levi’s, the Facebook Like buttons are not only problematic from a pure performance perspective, but from a general user experience perspective as well:

  • Every little detail on product pages can have an impact. While Facebook Like buttons might be cool, and might even be a good promotional tool, Levi’s extensive use of them on product pages is questionable. They’re given a very prominent position even though a significant number of consumers might not know what they are, and when clicked (either intentionally or unintentionally), it interrupts the shopping experience.
  • They’re a privacy and security risk. Although mainstream consumers might not be aware of the privacy implications of the Like buttons, retailers should be. Embedding any externally-hosted content on pages creates numerous privacy and security risks that cannot be controlled, and Levi’s use of them on retail-oriented pages shows, in my opinion, a certain disregard for the privacy and security of its customers and potential customers.

Of course, it wouldn’t be fair to ignore the potential benefits of the Like button. According to Compete, received approximately 451,000 unique visitors in the U.S. in April, and Facebook was the biggest source of referral traffic, driving 15.83% of it (Google, for comparison purposes, drove 15.66% of referral traffic).

Obviously, only Levi’s knows how Facebook referrals convert as compared to other traffic sources, which is really the most important metric, but what we do know: Levi’s has been investing significantly in Facebook lately, so this isn’t ‘free‘ traffic by any stretch of the imagination. And given that most products seem to have no more than a few hundred Likes at most, it is safe to say that Likes are unlikely to be generating the majority of the Facebook referral traffic.

Additionally, it should be pointed out that referral traffic is something that can largely be controlled. For instance, while Levi’s does seem to be using Google AdWords, using a U.S.-based proxy I don’t see much evidence that it’s targeting long-tail keywords, or using keyword-specific landing pages. For instance, a search for “Levi’s 510” returns a generic Levi’s AdWords ad that leads to the homepage — not exactly the ad a savvy search marketer would create to optimize for clicks and conversions, especially when other retailers, such as, are advertising against Levi’s-related keywords with keyword-specific copy. Bottom line: while Facebook is no doubt driving traffic to, that’s likely a result of the company’s broader investment in Facebook, and it could just as easily make its paid search campaigns more prominent and effective.

At the end of the day, even if one accepts for argument’s sake that Facebook Like buttons are wonderful, the fact still remains: the non-negligible performance hit is easy to establish. And in today’s online retail environment, the costs of these performance hits shouldn’t be underestimated.