by John Sherrod, Direc­tor of SEO

To redi­rect or not to redi­rect? That is the ques­tion.

Even the most orderly and well-kept dynamic websites can inad­ver­tently create a vast digital foot­print of unde­sired pages (unique URLs) that can poten­tially clutter Google’s index. Test pages, error pages, redun­dant, retired and unde­sired pages; pages that shouldn’t exist, and pages that don’t exist (yet somehow do) can jumble a domain’s pres­ence in search engine data­bases.

It’s the equiv­a­lent of a person clut­ter­ing up their closet with an outdated wardrobe. It’s time to throw out those para­chute pants already. And the strong-shoul­dered jacket from the 80s can go too.

Bing is a bit more exclu­sive about what gets in their index in the first place, but Google grabs as many URLs asso­ci­ated with a site as they can. And in Google’s case, all pages on a domain affect the site’s overall percep­tion of rele­vance. So it is impor­tant that webmas­ters make every effort to de-clutter their website.

Some quotes from a Googler (Wysz) clar­i­fies some of the char­ac­ter­is­tics of shallow content that was targeted by the Panda updates:
“…it’s impor­tant for webmas­ters to know that low quality content on part of a site can impact a site’s ranking as a whole.”
“…Bear in mind that people search­ing on Google typi­cally don’t want to see shallow or poorly written content…”
“…Remov­ing low quality pages or moving them to a differ­ent domain could help your rank­ings for the higher quality content.”
Read the whole thread here.

So, yes, it’s impor­tant to keep your domain free and clear of thin pages that add no value. But there are many ways to do this and not all scenar­ios are the same.

There are three main choices when it comes to address­ing unde­sired URLs on a domain.

  1. Delete them and just let them fail (404 not found error)
    1. Redi­rect those pages (sure, but which ones?)
    2. Leave them alone (just do nothing – Unde­sired!)

Some pages need to fail. They just do. Review­ing the advice from Wysz above, remov­ing / delet­ing (not redi­rect­ing) low quality pages is some­times the best choice.

Page types that should be deleted:

  • Test pages, “hello world” pages that other­wise have no content and odd pages that were never meant to be viewed by an end user
  • Error pages that are the result of a soft 404 (ex. an error page that redi­rects to an addi­tional URL that uses the broken query as a URL para­me­ter such as /Error404?aspxerrorpath=/supercalifragilisticexpialidocious
  • Pages that shouldn’t exist AND have never existed in the past.

There are certainly many others (and please add your sugges­tions to the comments) but I’ll move on to pages that should actu­ally be redi­rected.

In a nutshell, redi­rects should be employed when pages (URLs) have accrued value. Value can be inter­preted has having pageR­ank, being the target of inbound links, or having signif­i­cant age.

Let’s also be clear that any page that is being redi­rected to harvest legacy value should be a 301, perma­nent redi­rect. If you need to test your redi­rects this is a good source.

Page types that should be redi­rected:

  • Pages that have accrued pageR­ank. Use a chrome plugin to test.
  • Pages that receive traffic. Take a look at Analyt­ics reports.
  • Pages that have inbound links point­ing to them. There are lots of tools to test inbound links. Check GWT – Google Webmas­ter Tools ‑or use a tool like ahrefs.
  • Pages that have history. The major­ity of instances when a page with age doesn’t have pageR­ank or inbound links occur during site redesigns. URLs may change exten­sions and/or page names and it is impor­tant that redi­rects occur 1:1 to the new file loca­tion.
  • Product pages that have been removed or no longer sold / avail­able.
  • When pages have Canon­i­cal differ­ences. For example, the non-www and www version of a page both resolve. A 301 redi­rect should be employed for all pages so that only one version resolves.

What crite­ria to you use when decid­ing which URLs to delete and which to redi­rect?