Need help technically with dealing with "no results" pages on our internal search engine.

[Ray-pp]

Hi ntcma,

Generally, you can let Google naturally drop 404 errors from their index. You should not see any negative impact from the 404 pages errors.

However, if you're users reaching many 404ed pages, that would warrant a close look into the usability of the site.

• Why are users reaching the 404 page?

• What can you do to direct those users to the correct area of the site. For example: a message like, "Oops, looks like you searched for XXX, but that isn't on our site. Were you looking for YYY ?"

[ntcma]

Hi, I appreciate the overview

I am more looking for help with the technical behaviour of our current setup, I suspect it is suboptimal (but perhaps Google can figure it out just fine).

I edited my original post to try and make my request more clear.

[Ray-pp]

I would be concerned that your internal search pages could raise a duplicate content issue.

Currently, when searching on your site the results page has a canonical tag to itself. This would tell Google to crawl and index all your search results pages. Google has stated that they do not want to send users from their search to another search page, they would rather send users to a product page or content page.

I suggest thinking of a better canonical page to use for your internal searched. I.e. the main search page.

For 404s, those can be ignored and Google will drop them accordingly.

In Google Webmaster Tools you can tell Google not to crawl/index site URLs with certain parameters. I.e. you exclude your search parameters (and pages).

Lastly, canonical tags and GWT parameter settings are a suggestions for Google, which they do not always follow. If needed, apply rules in your robots.txt file to ignore the areas of the site needed.

[ntcma]

Thanks for the detailed consideration!

Again though, I'm looking for some advice on the technical working of our 404 rules, I'm concerned that we're not getting things quite right.

For example, when using a header code checking tool on the above page, I'm getting some mixed results (depending on the tool)

[Ray-pp]

I see a 302 redirection to a 404 page.

A 302 redirection would concern me - it tells Google that the page existed and is being temporarily redirected to the 404 page. That may cause Google to hang onto the url in their index.

I would adjust so that any 404 page returns a 404 error without redirection.

The exception is if you have a page where there is a lot of authority being passed to it and you do not want to lose that authority. In that scenario you would redirect the page to its closest relative (the page you want to have the authority). However, I don't see that being a concern with your search results pages.

[ntcma]

We need to compete with these aggregate pages, our competitors are ranking excellently with their aggregate results (i.e., search results) pages.

Our individual listing pages rank well and we receive a lot of inbound traffic to them; it's these SRPs that we're having trouble with.

We have taken steps to try and ensure responsible rel="canonical" usage. If you notice a case where things don't look right, please share!

Thanks for the reminder about the URL parameters in GWMT, we do use it, but it's been a while since I reviewed this

[evolvingSEO]

Hi There

I read the questions and discussion, but not sure I fully understand the question. Maybe you can clarify a few things?

• How does one get to this URL to begin with --> http://www.rvt.com/New-and-Used-Triple-E-Class-C-RVs-For-Sale-On-RVT.com/results?type=Class%20C&state=California&manu=Triple+E&searchtype=search
• Are you saying it IS an issue that it redirects to this? --> http://www.rvt.com/404.php or do you want it to redirect to that?

[ntcma]

Hi,

• One might find this URL in the Google cache for example, but on arriving to the site find that the current page points to no inventory, which at this point we treat as a page not found error.
• Yes, because it looks like a 302 from what I can see, which then issues the 404 ... seems like a less than ideal way to issue a 404 for that URL (if that's what we should be doing at all)
• An alternative is to actually issue a 200 and print a message like, "No results today, but come back soon or broaden your search". The big problem I see with this is controlling what are actually legitimate pages vs. nonsense pages (we should have some legit 404s in other words)

Thanks for the response

[evolvingSEO]

Hi There

How does it get in Google's cache? Is there a crawlable link to it somewhere internal on your site? Is there a way the user can get to this link? In other words, how is this URL created to begin with?

[ntcma]

Hi, ok, turns that was a mysterious example!

But in general, I suppose that one day we might have results, which yes are reachable through a link for Google to find.

The next day the link may be gone if we have no live results, but Google will not have realized that it's a 404 yet until they go back to it themselves (since the link is gone)

In the meantime though it is a 404

[evolvingSEO]

Gotcha - so are you all set for now? FYI, Google will re-crawl 404s in their index even if they are no longer linked to.