Back in December 2003 following the infamous "Florida Update" by Google, it was reported in this post that
Usually when a site is ranking high in Google for a very competitive keyword I can have a look at it and figure out why. Often it is because they have bought a bunch of incoming links or they have exchanged links with many other websites and their own links pages are either not listed or their links can't be followed which makes search engines thinks they have tons more incoming links than outgoing links.

There have been some sites that have stymied me though. They consistantly rank within the top few sites and examination shows that they have traded very many links over the years (often hundreds or thousands).

They seem to be on the "up and up" and their links pages are generally listed in Google and their links should be able to be followed (i.e. written in html and not scripting and they are not using "nofollow" or disallowing in robots.txt)

One recommendation that you will hear often in SEO forums is to validate your html. Last night I happened to be checking out one of these sites and found that the links pages had serious problems with overlapping tags that probably rendered them useless - all the [a href] tags were spanning tables, paragraphs, etc and even though the pages rendered properly in a modern browser I have no doubt that search engine spiders would be stymied.

When validating html is it common to receive error messages but there are some errors that cause more trouble than others - multiple overlapping tags, multiple tags that are not closed, missing quotation marks and brackets, and tables that have problems with their table row and table cell tags being the worst.

After viewing that first site I decided to check out a few others and looked at 4 more sites that also ranked well and seemed on the up and up.

Guess what? I was 5 for 5 with finding serious problems with links pages!

The most common problem was with tables - links were placed in tables and the tables were likely not being spidered because of problems in the code. One page reported over 2,000 errors. Another had code problems so bad it even prevented the source code from being viewed - you know for sure that those links are not being counted! One of the sites did not have their links pages spidered (url only in Google's index) and another was using a doc type that did not allow for tables - so the links would not be spidered.

So, I guess we should all dig a little deeper before approving links exchanges - maybe especially with sites that brag about their high Page Rank and positions.

FYI the sites I checked out were either the top ranked site or top ranked Realtor's site in the following areas....

Orlando
Orange County
Las Vegas
Edmonton
Calgary

Doug,

Great article, I'm sure that many of the members will find this a good read.

Another thing I just want to add to your article is "Check Their Cached Pages".

What some of these guys are doing is when their server side script noticed that Googlebot has come by to cache their page the either place "nofollow" on all of the links or they simply take all the links away.

If that happens then you get no credit for that backlink.

My 2 cents

Lots of interesting theories. Consider this, Google, Visa, Walmart, and a host of others, knows what society is buying today, searching for today, etc.

Goolge is no more than another looking to provide a service that is in demand. A tag here, a link there, some of the technical necessities. Who gets ranked depends on how well they match the desire.

Judging a page by the cache isnt correct. There are several valid reasons to not have your pages cached/archived by Google. Simply adding "no archive" in your robots.txt will do this.

This does not render the page unspiderable, nor does it mean it isnt in Google's index. It also doesnt mean that links on these pages are not crawlable.