« Article Bot is fine | Home | Personal Area Network… »

The value of 404 incoming links

A few thoughts that together will make sense in a second:

1. Google does not like "Moved Permanently" 302 (301?  I get those mixed up!) redirects because they have been overused by blackhat SEOs
2. links to http://example.com/index.php, http://www.example.com/ and http://example.com/ are all different, and all accumulate PR (and other ranking from other SEs) seperately.  Ideally you want all the links to go to the same place here.
3. You do want people to update links to your site if they are wrong - ie 404s, and if they have linked to http://example.com/index.php when you want to link - and pass PR to - http://example.com
4. 404s to a spider will through away any accumulated PR for that page
5. A page with a single link on it to another page will pass all it's PR vote to the page it is linked to
6. This PR should be used - passed to the site index for example and not wasted.
7. You want people to stop linking to the various flavours of your main site index - but get value out of them even if they never get changed
8. I HATE that I this blog for example has a PR3 for http://www.frakkle.com and a PR3 for http://frakkle.com.  If all the links went tohttp://frakkle.com then maybe this site would be a PR4...

Okay, I'm not going to give the details - though it would not be hard - I probably will when I get around to implementing it however.  How about the following:

Using this site as an example, assuming I want all links to go to http://frakkle.com/
Use spider detection, and apache's mod_rewrite so that:
For http://www.frakkle.com/ANYTHING

• Spiders get 1-link page to http://frakkle.com/ANYTHING
• Users get 302 redirect to http://frakkle.com/ANYTHING

For http://www.frakkle.com/index.php

• Spiders get 1-link page to http://frakkle.com/
• Users get 302 redirect to http://frakkle.com/

For bad links to pages that do not exist (404s)

• Spiders get link to http://frakkle.com/
• Users get 404 error and the same link to http://frakkle.com/

Maybe that last bit isn't such a good idea - might leave all those "dead" pages in the index for ages.  The first parts are clean however.

---

No comments:

---

No trackbacks:

Trackback link:

---

Name:			Remember personal info? Yes No
Email:
URL:
Comment:		/ Textile
this is to stop spam bots causing me pain. the first four letters of frakkle     (Register your username / Log in)
Notify: Yes, send me email when someone replies. Hide email: Yes, hide my email address.
Small print: All html tags except <b> and <i> will be removed from your comment. You can make links by just typing the url or mail-address.