Text Blog Software

When it comes to claims of effective SEO techniques, I'm a skeptics skeptic. In my line of work, I read and hear a lot of claims about what works.

Most of the reliable advice is just plain common sense. There's also a pool of good advice that probably makes sense only to those who work extensively with HTML and algorithms.

On the other side of the quality spectrum you'll find advice that falls into two bins:

shady advice that attempts to game the system. It might work for a while, but once the search engines adjust, rankings will plummet. Black hat SEO professionals make their living dishing out this dreck.
harmless, ineffective advice that buys you little or no benefits. This class of advice is usually rooted in empirical evidence that has been poorly analyzed and then passed around by word of mouth.

Last week Chris Boggs wrote a great post over at SearchEngineWatch.com that reminds the user that SEO advice requires critical thinking skills.

Would you make a life or death decision about caring for a sick infant based solely on blog advice, or even WebMD? Would you decide on a retirement strategy by reading a few blog posts or a forum discussion? Probably not. Yet, some marketers choose to base their SEO or paid search strategies on blog posts.

A person without any SEM experience shouldn't trust any single online authority as guidance. Misinformation and outdated tactical guidance seems to be floating up, along with the reliability of many so-called "SEO tools."

He goes on to write:

Design and development teams are usually great at sniffing out immediate problems with tactics, but also sometimes need to be reasoned with in order to test a theory or new SEO-friendly workaround. I've received plenty of blog post links from developers and designers in the past, claiming something can't be done or doesn't work. It's crucial to reach an agreement by either finding evidence or testing that the tactic can work.

He's right. If you're going to come at our team with a request to change something about a page for SEO benefits, you better have some hard data with good analysis to go with it. The whole post is worth a read and your time.

I thought of this post as I read a listing of tips on image SEO written by Patricio Robles. I found myself disagreeing with some of Robles' arguments.

Take the "descriptive name for an image" advice. Providing descriptive label helps, but I would speculate that search engines don't give images nearly the level of weighting that Robles claims. Unlike text, which can be processed and characterized by an algorithm, images can't be dumped into a machine and then identified. The means that a someone seeking to game the system could give irrelevant images names of keywords they are trying to target.

I don't buy the "larger images" advantage, either, which the author admits there is no hard data to back up the claim. If anything, larger images will slow down the rate at which pages load, making your visitors bounce very quickly. When we shipped a new editor interface a month or so ago, we even added new functionality that resizes images automatically if they exceed a threshold size that would exceed the typical width of a column. This was to counteract something that Randy, our UI Engineer, refers to as dumbnails -- large size graphics that are displayed at sized much smaller via inline styling on the image tag.

The advice to those who are trying to optimize their website? Beware and be wise!