A reader from Japan kindly points out that I have a subtle gaffe on page 45 of the book, where I talk about the link: attribute to a Google search and encourage readers to experiment by searching for link:askdavetaylor.com.

Of course, if you try that (link:askdavetaylor.com search on Google) you won't get the results I promise, but instead will find out that there are no matches at all!

Fortunately, that doesn't mean that I'm no longer in the Google database nor that there are zero sites that point to my site...

To see that there's no problem, do the search again, but this time use the pattern link:www.askdavetaylor.com. (link:www.askdavetaylor.com search on Google).

Surprised? Over 1100 matches.

The difference is because in this instance, Google interprets the search pattern quite literally, and when you ask for sites that link to Ask Dave Taylor by using a URL that contains askdavetaylor.com it finds zero matches.

You might be thinking "Ah, but surely www.askdavetaylor.com is a legit match, isn't it?" but in this instance it isn't. With some other special keyword searches, like site:askdavetaylor.com you'll find that the alternative of site:www.askdavetaylor.com gives you identical results.

There are some important implications to this link: sensitivity, actually, the most important of which is that you don't want some people pointing to your pages using XX.com while others use www.XX.com. If you do have these mixed, then Google can't properly calculate the inbound links for a specific page and will not calculate a fair and accurate PageRank for that page.

Here's what I mean.

Let's say that you really liked my article on ADT about how to find out your Google PageRank on a Macintosh (the article's titled Google PageRank on a Mac?) and you point to it by using the correct URL:

http://www.askdavetaylor.com/google_pagerank_on_a_mac.html

Meanwhile, your pal three cubicles over also likes it, but she points to it as:

http://askdavetaylor.com/google_pagerank_on_a_mac.html

This means that each URL will get credit within the Google algorithm for one inbound link. Not two for the page. Multiply this by a few dozen sites and you can see where this can actually be a problem.

My advice: if you have matches both with and without www., then email some of the webmasters for sites that have the wrong link and ask them to fix it for you.

Oh, and my apologies for the mistake in the book!