SEO and Validation

This post is based on a thread (Re: My Google Page Rank is 4) in the NG alt.internet.search-engines and a request for help in the comments at SEO Expert page.

Original poster asked about his 3-4 month old site www.cre8ivemind.com, with a PR4 home page, but no pages indexed in Google despite using Google site map.

Confirmed with the Google toolbar the home page was indeed PR4.

Click for large PR4 screenshot

Confirmed with a link:http://www.domain.com/ search in Google that the site had backlinks (5 backlinks) did the same with Yahoo, though Yahoo has changed it’s link: format recently and sometimes it doesn’t work (used to work with any site including deleted ones!)! Yahoo showed 113 backlinks.

Click for large Backlinks screenshot

Confirmed with a site:http://www.domain.com/ search in Google that the site had no pages indexed in Google! Did the same with Yahoo, Yahoo showed 8 pages indexed.

Click for large Site Search screenshot

So the site has a PR4 home page, backlinks, but no pages are indexed in Google. 113 backlinks on Yahoo should be more than enough to index this site, so as an expereinced SEO consultant I know something is wrong.

A very quick check of the home pages code (view source with Internet Explorer or View Page Source with FireFox) didn’t immediately show any problems. The site was created using the WYSIWYG (What You See Is What You Get) editor Frontpage and so there’s a lot of extra code (makes it harder to check manually for serious code errors).

On this first manual check I saw Flash code which matched up with the sites navigation (bad for spidering/SEO), but also found corresponding text links to the same internal pages, so if there are no major code errors this site should be spiderable, also found this meta tag-

This is telling all search engines to index this page and follow all links, BUT this is also the default for all search engines (they do by default follow and index), so pointless adding it. You should check this though just in case you added an exclusion like noindex or nofollow, see more Robots meta tag options at Meta Tags Optimization

It’s also a good idea to check for a robots.txt file by looking at http://www.domain.com/robots.txt This site didn’t have a robots.txt file, so this can’t be a source of the problem.

I couldn’t see anything else obviously wrong with the code, but it’s messy code so not easy to follow. This is when I use a validator specifically the W3C Validator.

Result: Failed validation, 45 errors

Anyone not familiar with validation may think 45 validator errors sounds bad. The vast majority of pages online today (August 2006) have numerous validator errors, yet work fine in the main browsers and the search engines can index them (check the top 10 ranking sites for almost any SERP, most pages won’t validate). Basically the browsers and the search engines can handle many validator errors, but they can’t handle ALL code errors, so we must remove the serious ones you find manually or via a validator.

We could try to remove all validator errors so we are 100% certain the code is serious error free.

Most of the validator errors reported are like these

No DOCTYPE found! Attempting validation with HTML 4.01 Transitional.
there is no attribute “LEFTMARGIN”.
there is no attribute “HEIGHT”.
required attribute “ALT” not specified.
value of attribute “ALIGN” cannot be “”; must be one of “TOP”, “MIDDLE”, “BOTTOM”, “LEFT”, “RIGHT”.
element “EMBED” undefined.

they are not going to cause problems, generally anything mentioning an attribute isn’t going to cause spidering problems.

There was however two validator errors involving the head and body tags that concerned me. These are important tags and to have an error associated with them is serious.

Line 22 column 6: end tag for element “HEAD” which is not open.
Line 23 column 59: document type does not allow element “BODY” here.

So if you ever find an error like the above two investigate and fix.

Looking at the code again I find a complete set of both tags (one each for opening the head and body tags and one each closing the tags) both in the right place. Hmm weird, I expected to find an extra closing head and body tag!

I then noticed just above the head closing tag some incorrectly formatted code for this Doctype (1st line below)-

<META name=”verify-v1″ content=”YG0wFdONXpRc175v3+BUnh67nykv+vqzaH/H5B1z27o=” />
</head>
<BODY LEFTMARGIN=0 TOPMARGIN=0 MARGINWIDTH=0 MARGINHEIGHT=0>

We use / there when using XHTML (which this page clearly is not), so +vqzaH/H5B1z27o=” /> should be +vqzaH/H5B1z27o=”> since it’s not the right format for this Doc type (HTML 4.01 Transitional is what the validator is working with)

I’m not familiar with that meta tag, so I’d delete (along with 90% of the other ones used!).

I don’t think this would cause problems with spidering, but best to fix it (could be the cause, who knows for sure how Google treats this sort of thing). By fixing this one error both the body and the head validator errors will be eliminated.

The head/body errors though potentially devastating I have a feeling Google would be able to handle them (several browsers versions/types handle the code fine, as does Yahoo which indexed 8 pages), so I think the problem is in another area.

There’s one other (set of) potentially serious validation errors-

Line 90 column 48: document type does not allow element “A” here.
Line 90 column 85: document type does not allow element “A” here.
Line 90 column 146: document type does not allow element “A” here.
…

with this code-

<a href=”http://www.cre8ivemind.com”><br>
<a href=”http://www.cre8ivemind.com”><a href=”http://www.cre8ivemind.com”>Home</a> l <a href=”http://www.cre8ivemind.com/about-us.htm”>About-us</a> l

The first bit-

<a href=”http://www.cre8ivemind.com”><br>
<a href=”http://www.cre8ivemind.com”>

Shouldn’t be there (causes over a dozen potentially serious validator errors).

The author has probably been dragging and dropping links and/or deleting them in a WYSIWYG type editor, ahh, Frontpage :-).

Sometimes when you delete a link in the WYSIWYG editor it only deletes the anchor text and not the full link. A little dragging and dropping later and you get the sort of code mess above. Not sure how Google etc… will handle this, so not the sort of error you want to leave to chance. This is the sort of error that could cause spidering problems, though shouldn’t prevent indexing of the home page (so not THE cause).

There is nothing else wrong with the code from a spidering perspective, so if the above isn’t the cause of the problem next is to look at possible penalties.

Didn’t check for penalties, IE duplicate content problems etc… since this post was about validation and SEO. That said the content isn’t very good, it’s mostly a list of keywords and so not of great importance. So might be we are seeing a site penalized for keyword stuffing. If it is that I’m impressed Google caught it (assuming it’s an automated penalty).

When I get some time I’ll take another look to see if I can find the source of the problem.