No other search criteria, just yoursite.com? Hm. Largest proportion of non-text, or of worthless text? That wasn't aimed at you: I just tried my own sitename. #1 was the index page-- which nobody ever goes to except when they're randomly exploring after landing elsewhere. Nos. 2 and 3 are reasonably popular pages. No. 4 is a page that nobody goes to; it's a subdirectory index page leading to-- among other things-- one of the most popular pages, but never visited in its own right. And so on.

Definitely no relationship to the listed pages' overall frequency in search results, or the number of sitewide keywords, or click-throughs from google.

Answer #1: It's something so embarrassingly obvious we will both be kicking ourselves.

Answer #2: It's some arcane, mysterious and hitherto unsuspected component of the g### algorithm.

I think the site command is broken. My two newest sites show the home page at #8 and #10 respectively, and otherwise are doing just fine.

This can actually be the sign of a potential issue, or it can be the classic Google bug.

None of my sites show this now, but have in the past. Most of it was a google glitch, but a few had real, tangible canonical issues that needed to be addressed.

Using example.com to represent the domain name in question, what is the displayed URL for the home page?

In my case, it just says example.com/

The page titles and meta descriptions are exactly right, just one is (today) at #4 and the other at #7.

Are all pages of the site indexed as non-www?

Do you have a site-wide www to non-www 301 redirect in place.

Does one of these searches return zero results, or do they both return some results?

site:example.com -inurl:www

site:www.example.com

Does it indicate an SEO problem when a site's home page isn't at or even near the top of the results for a site:domain.com command? If so, do we have any ideas what particular problem(s)?

Canonicalisation problems, internal linking problems, duplicate content problems, and others.

Argh, damn you glsmd, you made me go double check and sure enough I have a www issue on those two sites, which I totally missed because the home page is fine. So now I have to eat crow - in my case it might not be a bug, it could be because something was out of order in .htaccess.

Moral of the story is, don't take anything for granted.

thanks!

Oh, wow! I just trotted out the usual "stuff to go check", and didn't actually notice who was asking the question. :)

I nearly messed up big time on the .htaccess file for a MediaWiki site a few months ago. I put two rules in the wrong order, missed a rule out and had a typo in one another. Good job I checked the site logs early the next day; averted a potential disaster. Google had only picked up 3 duff pages by that time. Even so, it took Google more than a month to undo the problem. Doesn't bear thinking about what would have happened if I hadn't looked back for several weeks - or months.

Well fortunately I'm off season for these sites; once I took the three minutes to fix 'em, they'll be fine in due time. Geezopete.

Does one of these searches return zero results, or do they both return some results?
site:example.com -inurl:www
site:www.example.com

-inurl:www returns no results in my case so It's not a www vs non www issue. I'm not seeing any canonical issues either. What's odd is that in my case the #1 is my newest page and the #2 is my oldest page which happens to be the terms of use page? What are you telling me Google?

This kind of site: oddity can mysteriously appear from time to time and then just as mysteriously disappear. It is worth watching, but if your traffic and rankings don't change at the same time that the site: operator results go buggy, it may just be that mysterious "something" that happens. Google's back end data can get very weird when it comes to reporting functions as opposed to real keyword search results.

That said, sometimes I've seen it show up at the same time as what appears to be an algorithmic spam demotion. In that case, it's the algorithmic hit that matters most, and the site: operator results are a secondary sign.

I wish I could say it's cut and dried... but I've seen a lot of strangeness in this area and can't pretend to have it 100% figured out.

I had this problem about a week ago. My index page is optimized and ranks for the best keywords. When someone pointed out to me that my site didn't rank anymore, I noticed that all but my index page were stil ranking. For some reason my index page got de-listed completely (not in site: either) and after I sent through a reconsideration request, Google responded that they did not see anything wrong with the site (and then the index page mysteriously returned).

Must be a bug.

when I perform a site:example.com command for my site the index page is in position 3-5 beneath

This used to be a sign of the dreaded -50 penalty. Is it not anymore?

Even if -50 penalty is a thing of the past (haven't heard about it in a while) I would still check... Do you see a significant reduction in the number of long-tail KWs? It may not actually result in a big drop in traffic if you have enough of trophy KWs. It'll just look as if the traffic becomes "less diverse", so to speak. Helps to use log tools like Visitors [hping.org] that can show just the total amount of Google KWs. So, yeah, if the total number of KWs sending visitors from Google dropped, I would get worried...

I take a more pragmatic point of view: Index (home) pages are generally light on content, being mainly navigation, so having content rich pages appear above index/home is okay with me. Google doesn't display top/down, they display content-top/down...

edit: Bing, too...

I do personally believe this is a sign of some problem, as I have loads of site which return the home page once queried.

Hello all,

This is a common question. Having the site's root as the top result for a site: query is not guaranteed and nor is it a sign of a problem if it's not. It just so happens that in the vast majority of cases we return it as the first result.

Barry Schwartz of SE Roundtable blogged about it and his post generated a discussion on Google+: [plus.google.com...]

Hope this helps,
Pierre

Thanks, Pierre.

Matt Cutts and Vanessa Fox have said on multiple occasions that "site:www.example.com" searches will "in general" return "shortest URLs first".

will "in general" return "shortest URLs first"

I think that's where the suspicion of this being a sign of some trouble comes from. The root / homepage URL is always the shortest. If it doesn't show first, then the conditions are not "general" - always makes me think "what's wrong?"

"/" is a directory. "/index.*" is a url. "/a.*" will appear before index... and if using expresionless makes even more sense.

@g1smd:

My index page actually doesnt show up at all anymore.

Are all pages of the site indexed as non-www?

Is that the same what is set in webmastertools? Where you can configure display urls with "www"?

Do you have a site-wide www to non-www 301 redirect in place.

Im doing it the other way around, so if there's no www in the request I add it with a 301 redirect.

Does one of these searches return zero results, or do they both return some results?
site:example.com -inurl:www
site:www.example.com

Both return the same results.

What can be wrong?

[edited by: Robert_Charlton at 8:55 am (utc) on Jan 9, 2012]

If both searches return the same results, it looks like your canonicalisation redirect is not working or else it returns a status other than the required 301 (or you haven't waited long enough for the redirect to take effect).

Hmm, if I try to access example.com it redirects me to www.example.com, so that part seems to be working.
Also the redirect code is already in place for many months, it would be weird if only now Google would start to penalize me on it right?
Also its ONLY the indexpage that no longer shows up in the index, all other pages are still shown. So, I checked out this page: [google.com...]

But is there a way to see if Google has indexed my indexpage both as www.example.com and www.example.com/default.aspx?

Thanks! :)

[edited by: Robert_Charlton at 5:22 pm (utc) on Jan 9, 2012]
[edit reason] examplified domain [/edit]

[site:example.com inurl:default.aspx]

You should be redirecting requests for named index files to a URL ending in slash. This redirect should include the correct hostname in the target and the redirect should be listed before the non-www to www redirect. Failure to do that results in an unwanted multiple step redirection chain.