Site indexed w/o inbound links

John Scott · 10-16-2003, 08:27 AM

One site we are working on got indexed by Google without a single inbound link.

This is interesting because it has been the topic of some discussion whether Google indexes sites w/o inbound links.

PhilC · 10-16-2003, 08:30 AM

Did it ever have an inbound link - say, for testing purposes? How do you think Google might have found it - toolbar, domain registrations, searching in Google for the domain?

John Scott · 10-16-2003, 08:32 AM

Phil - sent you the URL. It has no inbound links. I'd be very careful not to link to it because it's not complete yet, and because it's an adult site. I don't want to link to adult sites.

It never had a link.

I can only imagine Google found it via toolbar.

PhilC · 10-16-2003, 08:37 AM

That's extremely interesting. It would be interesting to create a small, isolated section within a website but without any links to it, and visit it frequently. The toolbar always sends the url to Google to get the PR for the page being visited. That way, Google couldn't get the URLs from any domain registry, and the toolbar would be the only way they have of knowing about it.

John Scott · 10-16-2003, 08:38 AM

Brilliant idea. You want to do that?

Jazzee · 10-16-2003, 08:38 AM

I would be interested in the results of that.

PhilC · 10-16-2003, 09:03 AM

I'm being ganged up on

I'll set the test up later today, and I'll PM both of you to let you know the URL into the area - so's Google can't spot it. Then we can all visit it frequently. Alright?

Jazzee · 10-16-2003, 09:07 AM

Sounds good.

piramida · 10-17-2003, 03:19 AM

the way google can find your hidden site is trivial.

whenever you go to some other site having your 'hidden' site opened, it shows in referrer logs. whenever one of the referrer logs goes public, google can pick it up. webmaster of that other site might check it out and post links to some forum... you get the idea. i'd scan the raw logs for all referrers.

of course, the spying toolbar idea might be interesting

John Scott · 10-17-2003, 03:26 AM

Quote:

Originally Posted by piramida

whenever one of the referrer logs goes public, google can pick it up.

That assumes the referral logs are open. They are not. That assumes their are links to other sites - which would also be incorrect. With 4 different tracking scripts installed, if there was a link from another site, there's a good chance of it showing up in the referrer logs. But their are none.

piramida · 10-17-2003, 03:42 AM

no, that could've happend like this - if you have the 'secret' site open in your browser and then you type another location or a bookmark, the OTHER site logs the site that was open before it in the referral logs.

BTW, if your "4 tracking scripts" are external - they are a very good possible source of url leak, as well.

John Scott · 10-17-2003, 04:17 AM

Nope and nope.

Who uses external tracking scripts? That's sad. Very sad.

If Google found a page containing the link, a search would reurn it. Say, for example, the domain was hoho7789.com. You could do a search:
http://www.google.com/search?hl=en&l...ho7789+.com%22

Quote:

Your search - "hoho7789 .com" - did not match any documents.
No pages were found containing "hoho7789".

That's what I get when I search for pages containing the domain name.

Secondly, if there was a link to the site, it should show at least one referral from it. But no. The only referrer is http://216.239.57.104/search.

piramida · 10-17-2003, 05:08 AM

hm then it must really be the googlebar breaking your privacy. very weird indeed.

it has happened to many of my sites which I thought were developed 'secretly' and next thing I know the semi-finished website is spidered all over google.

so now I always use robots.txt with deny all or http authentification while in production, just to be safe. and the source of the leak in my cases was a. someones referral logs b. some external object linked from the page - affiliate button, banner, stuff like that. surprising really how, in one case, g picked up the uploaded web site *overnight*, i had it put under website.com/new for testing and in the morning I get website.com/new returning for the keywords i was optimizing for. had to get a redirect going from /new/ to / for a month since g was not dropping the broken url from the index.

John Scott · 10-17-2003, 05:11 AM

I am starting to use robots.txt on sites being developed. Seems to work.

Quote:

hm then it must really be the googlebar breaking your privacy

I think in order to have PageRank show in the toolbar, you have to turn "privacy" off. I could be wrong....

Mel · 10-17-2003, 08:53 PM

That is correct John, Google cannot show the PR of the page you are viewing unless it knows which page you are viewing.