|Can there be too many 301s and can they affect rankings?|
So, I'm in the process of rebuilding a large news section with an enormous amount of inherited problems. I've decided to make the urls nicer (this will also sort out a weird duplicate content issue).
It means several thousands of 301s. I spoke to this chap and he recalled a classified site having been dinged for introducing too many 301s. No-one can remember what site this was, so I cannot verify :)
1. Have you ever heard of a site that has suffered from having too many 301s? And I'm not talking about cases where 301 was implemented awkwardly. There are many ways you can get this wrong technically, but let's assume I do everything properly, i.e. old/ugly/article/url redirects to /new-and-shiny-url
2. Many of the articles are mighty old but I still want to keep them. If I change the urls now, when Google finds out the article resides on a new url, will it update its index or will it simply drop the old article? I've noticed that sometimes Google cannot be bothered re-indexing old articles but this was on a small and non-authoritative site.
3. A bit off-topic. If some of your news articles linked out to a relevant product page, considering the article was good quality and it really spoke about that particular product. Would you say this is a problem? Does that look spammy? From the visitor's perspective, there's nothing wrong linking to the Blue Widgets sales page from a news piece that speaks about a brand-new version of a Blue Widget. But since the ultimate goal is becoming an approved Google News source, I might reconsider what I do with links. What do you think?
1. Making major changes to a site often causes Google rankings and traffic to drop, at least for a few weeks.
2. A small loss of "pagerank juice" occurs when a 301 re-direct is used. This means that the entire site will be slightly weakened permanently.
3. There has been speculation that the Google algorithm gives less trust to sites that have large numbers re-directs, deleted pages, and other indications that a site has undergone a major structural revision at some point in the past. It could take years to rebuild any trust lost in this way.
A site previously with 80 000 duplicate content URLs now redirects and canonicalises those to 1500 product pages in a completely new URL format. There was a dip for a few weeks but several months later traffic is higher than it ever was. There was a long period where old URLs were still listed in the index, but some of those still brought traffic.
Google has said in a video I think that a 301 is no problem but when you redirect 3 times in a row it could give some troubles, but mostly related to crawling.
You can't tell from the thread title, but the OP made it sound as if the question is about number of redirected pages, not about number of redirect steps for an individual page.
Anyone have any figures on elapsed time? When googlebot meets a redirect it puts the second URL on its shopping list for later. How much later? Is it a predictable function of overall crawl frequency for the site? Does this second URL carry a tag that says "redirected"? What happens when the redirected URL meets yet another redirect? By the time your search engine bot arrives at a working URL, how much time has elapsed and how much extra baggage is the robot carrying?
It's hard to say "how long" as very often the new redirected-to URL is also added to the site navigation at the same time as the redirect is added. When the new URL is crawled, it's hard to say whether Google got the clue from the site navigation or the redirect. I just know that many of the new URLs can be crawled within 24 to 72 hours of the change.
|You can't tell from the thread title, but the OP made it sound as if the question is about number of redirected pages, not about number of redirect steps for an individual page. |
Just to make sure we're absolutely clear here... Matt Cutts confirms in the video noted below that there's no limit to the number of redirects Google is willing to crawl within a site, but there is a limit to the number of redirects you can chain together.
I've dealt with sites on which we redirected urls for hundreds of thousands of pages to canonicalize them and make them user friendly, and there were no problems and continue to be no problems as the sites have grown. (I should note that I did not write the code. Wise to leave that to a specialist, I feel.)
|I spoke to this chap and he recalled a classified site having been dinged for introducing too many 301s. |
It's hard to say, because of the vagueness of what the chap said, precisely what he himself understood or meant by "too many 301s", which, as we see, is open to multiple interpretations.
|Google has said in a video... |
This is the video....
Is there a limit to how many 301 (Permanent) redirects I can do on a site?
Matt Cutts - Aug 4, 2011
The questions, which Matt asks of himself (sounds like a put-up job to me ;) ) are...
|Is there a limit to how many 301 (Permanent) redirects I can do on a site? |
How about how many redirects I can chain together?
Quotes and rough paraphrases of some of what Matt says...
|No limit to the number of redirects... there's no queue; there's no cap; there's no limit.... We'll look at as many pages as we're willing to crawl on the old domain, and if we see a 301 redirect, then we'll put that in the queue to crawl at the new location, and we'll process that.... |
Matt encourages redirects page level to page level.... discourages chained redirects.... If you can do it in one hop, that's ideal. Google is willing to follow multiple hops... but if you start getting up into 4 or 5 range, that's a little bit dangerous, since Google might decide not to follow all those redirects. Keep it down to 1 or 2 or maybe 3.
Matt doesn't mention loss of PageRank per redirect in this video. We did discuss that in a March - June 2010 discussion here. The last post in the thread is probably where many of us have picked up the figure of a 15% drop in PR per redirect hop.
301 Redirect Means "Some Loss of PageRank" - says Mr Cutts
There have been a huge number of discussions in the Apache forum regarding how to combine redirects using mod_rewrite.
As I understand it, there may be differences in how Apache and IIS handle the logic of combining redirects, but I'll leave that for someone more knowledgable to describe.
My experience is similar to g1smd: In most cases when you have many redirects, it is owing to introducing new URL structure across the site.
What I saw is that in these cases Google spiders new URLs with a great speed - it is almost as it gives a priority to URLs that it has never seen to re-spidering existing URLs. The result is that for a while you may have two URLs in index: the new one and the old one that redirects, but the URL that redirects has not been spidered yet.
I have also seen old and new URLs swapping in ranking for some weeks (despite redirect) and then eventually settling on new URLs.
I am just about to put live URL restructure on a site that has 1500 pages, but where Google has indexed 8000 URLs + has further 16,000 as "Not selected" (owing to classic mistakes in parameter capitalisation, parameter order, extra parameters, dates in URL and so on). So we will be introducing 1500 new URLs, redirecting aproximately 1700 URLs and leaving other 22,000 go 404. The site itself is only 18 months old so it will be interesting to watch what happens, how long will fall last and how long to recover.
While Google is listing the old and the new URL, don't be tempted to manually remove the old URL from the index. Until the new URL ranks, it's the old one that is bringing the traffic. Let Google handle the re-indexing. If old URLs are still showing 6 months later then there's a problem.