1. Original Entry + Comments2. Write a Comment3. Preview Comment
New comments for this entry are disabled.

May 26, 2011  |  Blog update: Added a Facebook Like button  |  84341 hit(s)

Upon attending an SEO presentation this week, I learned that I should be enabling social media a bit better on this blog. So I added a Facebook Like button.

This is my first attempt, so I went with a button that shows up at the bottom of the entry. I was told that it's not bad idea to have a button at the top; a peculiar but attested behavior of users is that they'll sometimes "Like" something before they read it, or possibly without reading beyond the beginning. But let's see how this works out.

Implementing the button is not particularly difficult. The details are laid out on the Like Button page of the Facebook Developers site. In effect, what I did was add the following markup to the bottom of the layout for a blog entry (broken up here for readability):

<iframe src=
style="border:none; overflow:hidden; width:450px; height:30px;"

I had to add some code to the page to substitute the URL of an individual blog entry (e.g., http://www.mikepope.com/blog/DisplayBlog.aspx?permalink=2272) for %%1%% in the markup. (The blog code is something of a rat's nest anyway, so what's another 4 lines for this?)

You're also supposed to add some meta tags to the <head> element that Facebook uses (they say) to determine what title, URL, and graphic to associate with the content. For example:

<meta property="og:title" content="mike&#039;s web log" />
<meta property="og:type" content="blog" />
<meta property="og:image"
content="https://www.mikepope.com/blog/images/mikesblog_defaultimage.png" />
<meta property="og:site_name" content="mike&#039;s web log" />
One thing that's implicit in the way they implement this, and is noticeable to me already, is that there is a slight delay before the Like button is displayed. Each instance of the Like button has to call the Facebook site and get data about the blog entry that the button is bound to, then return a hunk of markup to display on the page. I display 10 entries per page; that's 10 calls to Facebook. (Only one call, tho, if you're looking at an individual entry.) On the plus side, this is done asynchronously -- not AJAX-y, just via a queued HTTP request, like images. That means that the actual blog content shows up as it always has, it's just the Like button itself that might show up slightly later.

Anyway, we'll see how this goes. If you have some feedback about this, please leave a comment. And, of course, be sure to Like this. :-)

mike   26 May 11 - 1:25 AM

Hmm. The image that's showing up on the FB wall is not matching what I specified, even tho FB's URL Linter (https://developers.facebook.com/tools/lint/) seems to show the correct image. Might be a cache problem, tho.

Jim Glass   26 May 11 - 7:06 AM

How are the Google ads working out for you?

Alan Humphrey   26 May 11 - 4:02 PM

I was experimenting with this for BirdWeb a month or so ago and noticed the same delay. Sometimes it was seconds before the button would show up (just one button on the page) other times it was obvious only if you knew where to look.

And then there's the boatload of code they download just to display the button, let alone support clicking on it. Surely this can be done more efficiently.

Anyway, the powers that be are evaluating whether we'll go forward with it. I suspect we will.

Anonymous   26 May 11 - 9:20 PM

I have talked to several people who went to the SEO thing. Not a single one seemed to fully understand what was going on there. At least, not well enough to explain to me. Apparently, during each explanation it showed on my face that it wasn't making any sense.

From my POV, SEO is a game that search engines don't want you to win. Kinda like the lottery. If everyone who creates content goes to a lot of trouble to try to win the game, then the people who really win are the people running the lottery. Yeah someone is going to win, but a lot more people are going to lose.

mike   26 May 11 - 10:42 PM

I think there was a, um, impedance mismatch between the presenters and some segment of the the audience (which included me). Almost all of the information pertained to engineering, not content. For example, I went to one presentation where they spent almost the whole time talking about the robots.txt file, which extremely few people have access to for the microsoft.com site.

I think that you're correct in that SEO for its own sake is not useful. In fact, they had a couple of counter-examples (ex: http://bit.ly/gCgPFb) of how big companies that tried to game the system were seriously punished by Google.

There was some generally useful takeaways. One was that good content builds organic links, hence rank. IOW, do what you've always tried to do, namely write stuff that people want to read. A kind of corollary here was that there are lots of tools that can help you understand the crawler and search-user POV; if you want your good content to be discovered, think like a crawler and like a user. A second key point was that social media is increasingly important for establishing "authority", and you need to make sure you're playing in that space and understand how it affects your discoverability. As one presenter put it, "social referrals are the new home page" -- that's how people get to your content.

If nothing else, I think I now know where I might start asking questions. That's something, anyway. :-)