Changing Page Names

[greenfrog]

Hello,

I have changed many page names on my site but left the original pages on the server, and simply added the new pages to the server. So essentially there are two exact copies of the same content, with different page names.

The new pages are linked, the old pages are not linked anywhere, but they have been left to catch any traffic sent their way.

I have seen google visit the site and view the new pages, I actually saw the pages begin to flow through the servers google-va, google-ex, google-in, and now they seem to have disappeared.

Do you think I have a penalty on my site? If google is following my links, they should not find the old pages or the old page names.

Any advice is appreciated.

[Made In Sheffield]

Don't worry about it, all will be ok in time.

Eventually the following should happen (in no particular order), time periods very hard to suggest at the minute:

1. Google will keep spidering the pages using Freshbot, the pages will be in one day and out the next. This is called EverFlux.

2. The pages will get spidered by the GoogleBot (freshbot also uses this name).

3. The pages will be added to the full index (2 has to happen first) and (nearly) always there.

4. The old pages will be recognised as no longer being linked to and dropped from the index (could take a very long time)

This is all based on the way Google has worked for a while, we don't know yet if this has changed so you'll have to keep your eyes on this forum to see if it's going to work differently in future.

Hope this helps,
Nigel

[steveb]

Either use a redirect to send people to the new pages or simply remove the old ones. Do not leave two pages of the same content up.

[g1smd]

At the very least put this on all of the old pages:

<meta name="robots" content="noindex,follow">

Allow 6 to 8 weeks for the old pages to drop out of the index.

[Made In Sheffield]

I have a colleague with old pages that have been there 6 months plus.

[pjnunn777]

In my experience Google will ignore one of the pages as it sees it (rightly) as duplicate content.

The problem with leaving old pages on the server with no links is that G will still ask for the old page as it "knows" it is there.

I found this continued for quite some time. It is much easier to put a noindex, etc and tell G that you do not want that page idexing.

[caveman]

Yes, if at all possible take the old pages down and add 301 redirects. Duplicate content can in some cases cause far more headaches for a few months than whatever traffic you will lose by taking down old pages.

With 301's you're ensuring that the SE's - eventually - will get the right pages indexed, and since the old pages will be gone, you're at least doing everything you can.

Post Dominic, Google in some cases seems to be having trouble wtih indexing changes like this, but as far as I know, 301's are still the way to go.

[marek]

OK, 301 redirect is certainly the best solution. However, any suggestion, what to do, if the web site is on a shared hosting, contains only static HTML pages and setting up the redirect is not possible?

[Herenvardo]

You can take out all the content of the old pages and replace it with something like this:

Sorry. By technicall reasons the URL of this page has changed to <a href="newurl">newurl</a>

You can modify the text as you need. For example, you can change URL with address if you visitors are not used to these terms.

It can be even better than redirect:
With a redirect, the user automatically reaches the new page as if nothing had happened. When s/he wants to enter the page again, s/he will use the same old address.
If you use the link, you are doing two tasks in one: you redirect both users and robots (and even PR) to the new page, but you are also informing your visitors about the change, so they will take note of the new url and use it the next times the enter your pages. They will even refresh their favorites to the new pages. After some time, google will drop the old pages and you will be able to delete them. Your common visitors and your inbound links will have had time to get refreshed to the new urls.
When only a part of the links have been redirected, the old links give PR to the old page and this, thru the link, passes it to the new one. Meanwhile, the already redirected links give the PR to the new pages. So you won't even lose PR during this "adaptation" process.

And, if you're not convinced yet, i'll give you another argument:
It is standard HTML and does not use any "modern" feature like frames or scripting, even doesn't use images!

Regards,
Hrenvardo

[greenfrog]

Thanks for all the great help.

I am running my site on a java Tomcat Server (Not an Apache Connector), and I can't find any info on setting up 301 redirect with Tomcat in the server.xml or web.xml file.

What I am thinking of doing is deleting the old pages, today, and setting up this to catch all of the missing pages.

<error-page>
<error-code>404</error-code>
<location>/index.jsp</location>
</error-page>

So, when a referral comes from google and the page is not found it will be caught by the 404 error and then redirect to the home page.

Is this a good solution?

[g1smd]

That is one of the three ways to do things. You may lose a few more visitors with a 404 though, as they may not be bothered to navigate the site for the new content. Make sure that you 404 page really does return a 404 code and not a 200 though.

I prefer the meta noindex tag (example above) if a 301 redirect is not available. I did that with a site at the beginning of May, and in mid-June Google suddenly listed all of the pages of the new site for the first time, and dropped all of the pages of the old site at the very same time. The old site had linked all "internal" naviagtion straight to the equivalent page on the new site.

[paulewing]

Just a word of warning, I did the page name change back early in the year and installed 301 redirects for all the pages. On the next refresh all the renamed pages went from PR5 and PR4 to PR0 and stayed that way for months. I have finally crawels back up to PR4 or PR3 on the pages, but it hurt my rankings for most of the spring. Things like dropping from #1 or #2 to #8 or worse for my admittedly only moderatly competative keywords and phrases.

Cheers,
Paul

[GrinninGordon]

Something paulewing said

There is some feeling amongst some webmasters that changing your directory structure and main page names is not a good idea. In that Google may not like changes in these, perhaps feeling people that change names are likely doing SEO, and perhaps reflecting one of the "tips of the day" given by w3c when you validate your html. It says something like "A cool URL does not change". Seriously, I was amazed w3c used an expression like that. But there you are.

During recent algo changes, some ancient sites that had not been touched rose dramtically during the dance. Which gave rise to this speculation. When I recently (December 2002) had to redo a site structure due to its expansion, it suffered quite a lot, even though I was carefull to put permanent redirects in the .htaccess file.

[ga_ga]

Yeah, 'Tip of the day: Cool URI's don't change' as I remember, might be wrong.. it threw me to see them put it that way too!

[tedster]

In the days when many crawler based search engines worked the web and drove traffic, there was more than one algo where longevity of the URL was a major factor. Could always come around again at any time.

People I work with sometimes think I'm crazy with the amount of time I spend planning a new site's directory structure and the entire name space I intend to use. I'm convinced that this planning pays off in many, many ways -- and having URLs that don't change is a big one.

[greenfrog]

I tried to catch the error code for java tomcat, but i have to work out a few bugs with the modular design of my site.

<error-page>
<error-code>404</error-code>
<location>/index.jsp</location>
</error-page>

I am including jstl tag libraries on my pages and this is somehow throwing a error code 500, which is frustrating. The error is weird and appears random, becuase many pages display fine, and others throw an error?

<%@ taglib prefix="c" uri="http://java.sun.com/jstl/core" %>

So anyway, I am going to go with a common suggestion in this thread and use the <meta name="robots" content="noindex,follow">.

I am still learning this sun, solaris, java, tomcat, and seo thing, so I appreciate all of your help.

[greenfrog]

Thanks for all the help, Googlebot spidered my pages within the last week, and just after I made the recommended changes in this post.

It read all the <meta name="robots" content="noindex,follow"> in the old page names.

It dropped them out of the index almost immediately!

My index isn't perfect because some pages are missing, but at least I'm on the right track.

Thanks for the advice.

[g1smd]

Nice! This happened to me in mid-June. It dropped every page of the old site and listed all of the new site for the first time the very same day.