Forum Moderators: phranque
To be less flippant, there is absolutely no evidence (that I have seen) that coding to W3C compliance is of any value for the search engines. Blantantly invalid code can fail to get indexed, and there are a huge number of benefits to correct coding, but making a validator absolutely happy has, IMHO, zero effect on the SERPs.
To be less flippant, there is absolutely no evidence (that I have seen) that coding to W3C compliance is of any value for the search engines. Blantantly invalid code can fail to get indexed, and there are a huge number of benefits to correct coding, but making a validator absolutely happy has, IMHO, zero effect on the SERPs.
I would hope that writing clean, W3 compliant code would help you in the SERPS.
I got a URL from someone here a while ago for a freeware tagchecker that works great. Let me know if you want it and I'll sticky it to you.
ADDED: Should have mentioned that I have the link on one page. I figure one link on the site to them is enough, and that page gets looked at often.
Even if the code is good enough without validation to get crawled successfully, (most cases probably), at least by validating you know that there aren't any major problems that might be messing it up so badly that a searchbot won't make it right through the page. Using a tagchecker really helps too if you're doing the pages by hand with a text editor. I always find mistakes the first time I check.
Writing clean, valid code ought to count for SOMETHING. :\
Writing clean, valid code ought to count for SOMETHING.
Short and sweet...
Two identical pages. One valid, one not valid. The valid one will outperform the invalid one 9 out of 10 times. That's enough reason for me. ;)
Two identical pages. One valid, one not valid. The valid one will outperform the invalid one 9 out of 10 times. That's enough reason for me. ;)
I very much doubt that a page using & in a URL will outperform one using &. Yet the latter is clearly and unambiguously invalid as far as W3C is concerned.
I am not arguing against validation, nor am I arguing in favor of sloppy code; I validate my pages. I am just suggesting that search engines are more interested in function (content) than in form (validity). So, perhaps, are users ;)
plenty of junk html looks ok in IE but screws up in netscape, opera, mozilla etc
if i were implementing a ranking algorithm i would certainly give credit for valid code, if you take the time to validate you've probably taken the time to create a good site.
To be less flippant, there is absolutely no evidence (that I have seen) that coding to W3C compliance is of any value for the search engines. Blantantly invalid code can fail to get indexed, and there are a huge number of benefits to correct coding, but making a validator absolutely happy has, IMHO, zero effect on the SERPs.
The problem is defining what is non-blantantly invalid code -- ie the code that is not correct to W3C standards but which is acceptable to all known spiders and browsers.
It's way too hard to guarantee to my clients that my code contains only non-blantant errors. It is much easier and safer to have no errors at all.
Woudl you hire a site designer that claimed "non-blantant errors a speciality?" Or one who wrote provably valid code?
Other background reading:
A HTML error that stopped spiders:
[webmasterworld.com...]
Wise advice from Brett:
[webmasterworld.com...]
... but making a validator absolutely happy has, IMHO, zero effect on the SERPs.
In a later post I gave a specific example of what I meant:
I very much doubt that a page using & in a URL will outperform one using &.
Anyone care to comment on that?
The issue in this thread is not whether validation is useful or nor, the issue is whether it helps with the SERPs. Let me repeat: my sites validate to 4.01 Transitional, and validate to Strict apart from the use of taget=_blank. I obviously believe that it is a good thing, otherwise I would not do it.
But I believe that it is incorrect to say that validation per se helps with search engines.
But I believe that it is incorrect to say that validation per se helps with search engines.
I gave you a counter-example where a validation error stopped a page being indexed.
It is likely that many errors are neutral when it comes to helping with most search engines.
It is also possible that some errors are positively beneficial for some engines (while being neutral or damaging with others).
It seems a prudent verify any error that someone is planning to add to HTML has no bad effects on any SE that they care about.
That's a hard task to do. Hardly seems worth the effort -- writing non-validated code requires an enormous research effort first.
search engines are more interested in function (content) than in form (validity).
Absolutely.
if i were implementing a ranking algorithm i would certainly give credit for valid code, if you take the time to validate you've probably taken the time to create a good site
Well, you wouldn't if you wanted to give users the best search results possible ;)
The problem is that you're assuming a validly-marked up site means the author has (additional) expertise in another HTML-unrelated domain (e.g. botany) - which imho is a pretty wild assumption...
Using validation in a ranking algorithum would skew results to favour pages written by the minority of webmasters/publishers who understand validation... but "valid" webmasters aren't neccessarily experts on anything other than valid markup, and domain experts aren't neccessarily also HTML validation experts.
I very much doubt that a page using & in a URL will outperform one using &. Yet the latter is clearly and unambiguously invalid as far as W3C is concerned.
Agreed. But, the example you provided is a minor issue. Let's talk about some of the other issues that could arise. One of the things I've seen stop a spider in its tracks...
Invalid HTML Comments
The entire page ended up getting commented due to a string of hyphens (------) inside the comment).
I'm sure there are many other errors that will prevent a spider from indexing a site properly. I don't know for sure as I've been writing valid code for over 2 years now.
I've probably run at least 50 sites through Brett's SIM Spider only to find that some of them were not allowing the spider to index the content properly. After cleaning up the html and correcting some errors, the spider was able to zip through the code and get what it came for.
I like to use a race car analogy in this case. Would you rather drive a race car that was tuned to properly mix air/fuel at every 10 rpm? (A site that has errors in its HTML/XHTML). Or, would you rather race a car with an engine management system controlling air/fuel at every 1 rpm? (A site that has no errors in its HTML/XHTML).
IMHO, validation does have an influence over a sites ability to perform in the SERPs, minor errors aside. It is so easy to make a mistake and create a fatal error in your code. Without validation, you'd never see it and may end up at a WebmasterWorld some day wondering why your site is not getting indexed properly. ;)
I very much doubt that a page using & in a URL will outperform one using &. Yet the latter is clearly and unambiguously invalid as far as W3C is concerned.
URL or HTML code?
if you take the time to validate you've probably taken the time to create a good site.
No it does not.
SEO is bad enough as it is, give them some more to work with why don't you.
SEO is bad enough as it is, give them some more to work with why don't you.
I bet.
