| This 43 message thread spans 2 pages: 43 (  2 ) > > || |
|Changing lots of file names at once = bad news|
Warning - Don't change all your file names at once!
In our naivety we thought we would be less spammy and changed our file names from cheap-blue-widgets.html to blue-widgets.html.
We did this throughout the site using redirects to let google know the pages had changed.
We then read (when its too late to change back) in the Google Guidlines that this can cause some havoc and ruin the sites PR and therefore ranking.
Well it has and now we only have 101 pages out of 8000 or so indexed and we rank nowhere - has anyone experienced this before and got any recommendations for getting out of this mess?
Its been 7 weeks and the site is being almost ignored.
6 or 7 weeks, nothing changes, then a few more pages come in as soon as I post on WebmasterWorld :)
still about 95% down on traffic, anyone been through this?
yes a few years ago i did the same, only a few page names but all the descriptions on all the pages got changed, the site went missing for 6 months until I changed them back. I was not a test it was a very big mistake and a large learning curve.
If it an't broken don't try and fix it
drastic changes have never been a good idea
I haven't been through it myself, but I've seen it happen.
Stick with whatever you got now, or you'll have the same experience ... again.
How often is this site crawled? Are you sure the redirects are 301?
/me hurridly changes some plans
I was under the impression that 301s would avoid this - I'm *sure* I would have researched the issue.. but, er, thanks for the warning LukeC..
I nuked my own site doing the same. Hope time heals all.
LukeC, did you use 301 redirects for your old files to your new files?
I think it doesn't matter that much whether it was a 301 or just nuking the old pages.
Both mean some quiet time ...
We want to do some work on our URLs:
What would be the best way to do a transition? Should I just keep both versions live? Delete the old one? Do a 301 redirect? Or never change a running system?
I'm sure no one else is silly enough to have followed my example, but before I renamed all the pages I uploaded them and tested them. To help avoid the issue of duplicate content during testing I added NOINDEX to my newly renamed pages.
Of course, after going live, I forgot about my NOINDEX. Once that little *bug* was corrected my new pages started showing up in a matter of weeks.
I've since learned to make significant changes like this in incremental steps, and to check, recheck and check my work again.
So I guess it's extremely risky in terms of staying in google's index to move an existing site to another domain with 301's, even if the 301's on the original site can be kept active indefinitely?
In this case the file names/structure/page content would all stay the same, but the actual domain would change...
Has anyone successfully done something like this recently?
about 10 days ago I changed most filenames on a site with about 1000 pages.
previously there had been (lots of) spaces in the urls which I changed over to be hyphens and used mod_rewrite to 301 multi%20word%20filename to multi-word-filename
within a day google had picked up the changes and crawled most of the old and new names. New names showed in the cache pretty quickly - it took me a few days to get round to it but all I've checked are there now.
gbot is still crawling (more slowly) but mostly has stopped trying old urls. slurp is crawling both. while poor msnbot is still frantically beating its head on the old ones.
I wasnt too worried about the effect on google rankings because this site got zapped by Florida anyway.
Pre Florida it was 1 - 5 on most relevant searches. Prior to the url changes it had slowly returned from oblivion to 20's and 30's (dont u just luv adwords)
Difficult to be conclusive after 10 days, but if anything, since the changes it has improved a little.
index page was/is pr5. a few remaining unaffected internal pages are still showing prev pr. any changed pages are now pr0. I'll post again when their pr returns (I think it will) or anything else changes.
What you did was start over again. It takes about 3 months to get that many pages indexed in G and that is if you have an old site with good pr that gets spidered a lot.
|6 or 7 weeks, nothing changes, then a few more pages come in as soon as I post on WebmasterWorld :) |
Talk about your mother of all conspiracy theories!
|within a day google had picked up the changes and crawled most of the old and new names. New names showed in the cache pretty quickly - it took me a few days to get round to it but all I've checked are there now. |
Exactly, your problem is that Google did not crawl your site as fast as you wanted. From the sounds of it you were expecting Google to just pick up the 301'd pages immediately? Like the last posted said, it will just view it as a completely new site and start from scratch.
I take it you didn't have good inbound links to those pages anyway otherwise you wouldn't have bothered changing the file names?
I guess we had about 100 inbounds, some of them must have been good as we ranked really well before the Aug 5th update.
Google has been visiting us regularly again and we now have 700 more pages in.
The view that its starting afresh still seems to hold as the rankings for the newly indexed pages are mediocre.
Yes we did do 301s, google is taking a while to get rid of the old pages. :(
For all those as inexperienced as me - careful when you change all your file names!
|So I guess it's extremely risky in terms of staying in google's index to move an existing site to another domain with 301's, even if the 301's on the original site can be kept active indefinitely? |
Yes, it's extremely risky, well, no, it's not risky, it's basically suicide at this point from what I'm seeing, at least for x number of months (currently coming up on 3), the google I used to respect for the their super quick adaptation to the web and websites is gone completely, 'freshest results' are things of fond memories. Google will spider your site fine, your pages will be indexed, but for some reason the current algo just can't deal with the fact of a simple domain name change, neither can yahoo from what I'm seeing. It almost looks like a misapplication of the duplicate content penalty to my eyes, search terms are buried deep in the SERPS after the domain name is changed and the old one 301'ed. But any new content is spidered fine, and comes exactly where the old content came up in the SERPS, so it's not really a sandbox per se, I think it's actually a mistake, a bug, since what possible purpose could it serve in any respect to fail to correctly promptly move a site in the index from old domain to new domain, when the old one uses a 301? It's almost but not quite sad, pathetic comes to mind as well...
But I'm not wasting any time trying to figure out the whys of this, it's not going to help get fresh results and quick updates back any time soon, I am waiting anxiously for the new MSN to come online however.
"I am waiting anxiously for the new MSN to come online however."
I'm not sure what you folk are recommending as a solution to the creation of a new website, Widgets Mark 2
Is the best solution - more importantly is it a solution at all? - to create a new website, Widgets Mark 2 as either a sub domain or just new pages to the existing site; and when these new pages are indexed and ranked to delete the old pages with the exception of the index page which can now be modified.
I make the proviso in this scenario that all the new pages are totally different in content to the old, so won't be penalised as duplicate content.
Is this a 'runner' or have I missed a trick - probably have - there is a lot to keep an eye on?
If I were making a new website and didn't have a previous domain name registered, I'd put up the new site and wait.
Or I'd buy a currently active domain name and build the site on that, as caveman advocated in another thread.
If I had a current site and was ok with the domain name, I'd just add to that site.
What I wouldn't do is anything that might involve having to 301 a subdomain or part of an existing domain to a new domain any time in the next year or two, or however long it's going to take google to once again start running a functioning algo and index.
Now if I were me, which I am, I'd just decide to see for myself, blow up my old, well established domain name, make a new, more marketable domain name, accept that google is not my friend for the next x months, and let it rest there.
Or get rid of that SE traffic and stick to "normal" traffic. When SEs pick up it'll be fine, if not it's all right too.
Isitreal - or anyone else able to help.
Obviously, I didn't make myself clear in my post.
My client has upgraded his software product - a product which he sells by mail order. The market is international and very competitive.
The current site is reasonably placed in the SERPS but way not good enough for the mail order business to really take off.
The client is pushing for a "better" site with which to tackle the SERPS. He needs a lot more interesting content ie marketing oriented pages to compete; and a lot more incoming links.
The idea it to do a makeover on the current domain, keep the file names as much as possible ( not sure yet how practical this is) and add new content.
However, the site is a FRAMESET -I'm not over happy with this.
QUESTIONS: Can I safely do away with the frameset without Google going beserk?
And, secondly do I have to retain browser titles and meta tags to retain page integrity?
And, thirdly is it an option to build a completely new site on the domain with totally new content ( so as to avoid the duplicate content penalty), put it up alongside the current site and when the new 'site' is ranked delete the old 'site' but keep the 'old index page' - preferably now without the frameset?
That's the idea. Will it work? Have I missed a trick?
How 'different' is different to avoid the duplicate content penalty?
I'm really worried I'll screw up on this redesign and lose the current position in the serps and make things worse rather than better.
Can you please give me your best shot at this one?
>However, the site is a FRAMESET -I'm not over happy with this.
Yup, dump the frameset concept and go with as static tables pages as you can. Try and use all the old file names, and a safe option is to try and maintain the old frameset linking pattern. In otherwords, if an existing frameset page links to 3 inframe pages, then create a new page with the old frameset filename, and do hard links from that page to new stand alone pages using the old inframe filenames. I reckon if you disturb the overall existing linking pattern too much you may cause problems.
>QUESTIONS: Can I safely do away with the frameset without Google going beserk?
Yes, but re above.
>And, secondly do I have to retain browser titles and meta tags to retain page integrity?
Nope, new titles could help!
>And, thirdly is it an option to build a completely new site on the domain with totally new content ( so as to avoid the duplicate content penalty), put it up alongside the current site and when the new 'site' is ranked delete the old 'site' but keep the 'old index page' - preferably now without the frameset?
That would be a simple option, but it could create a lot of problems. Doubling your site will effect and dilute pr flow, so overall rankings could be lost. There is also the duplicate content issue, which is harder to beat than many people think.
That's the idea. Will it work? Have I missed a trick?
The trick you have possibly missed is that your internal pages may have been indexed by hundreds of small directories. We have lost sites out of google, and still maintained good traffic from being listed in spammy little directories... don't underestimate their worth. Some give pr, some good anchor text, some good traffic. Deep pages may have been listed, providing on theme links. You may not easily find them, especially via link:google tool, but Yahoo etc. love them.
How 'different' is different to avoid the duplicate content penalty?
Big question! In googles attempt to reduce the number of directories ranking high, they have stepped this filter up many gears. I don't think they penalise for duplicate content, they just ignore it, unless it is fresh content or on a 'news' site (whole new topic... lets not go there!).
I think they look at small chunks, so beating the filter is very hard and not worth the effort. Having said that, the duplicate content filter is directed at different domains. Pages within a site may be a different matter. I would not duplicate pages within a site, even if you think they are different, they are duplicate whatever way you look at them. Try and keep the site the same size throughout the transition, there are too many potential hazzards by doubling your site overnight... it looks spammy and unnatural.
What an excellent post Mhes.
Thanks for taking the trouble to answer all my questions
Are there any other comments - pehaps dissenting with Mhes - to add to this one?
|The trick you have possibly missed is that your internal pages may have been indexed by hundreds of small directories. We have lost sites out of google, and still maintained good traffic from being listed in spammy little directories... don't underestimate their worth. Some give pr, some good anchor text, some good traffic. Deep pages may have been listed, providing on theme links. You may not easily find them, especially via link:google tool, but Yahoo etc. love them. |
My god you are so right, we changed our file names and now I realise how much they were worth, we had links to every page on our site.
Also it's like a snowball effect, the better positioning you have the more of these uselss directories link to you and then the higher you get positioned ad finitum...
OK! Agreed. Great care with existing file names.
But, why the aversion to FRAMESETS?
I've used two-frame framesets quite alot and have never been aware of any major disadvantage provided you use the 'no frames' content properly. I always put the navigation on the inside or content pages otherwise they can beome orphaned.
But, what other disadvantages are there?
>But, why the aversion to FRAMESETS?
Gut feeling is that you are asking a spider to index content (noframes) that isn't the actual content seen by 99.9% of users..... spam potential heaven, so google may be wary. Looking around in the past, it has been abused.
The other problem is that google index's the pages of the frameset individually, often making the search results list only part of a page which is often useless for the user. Smarter webmasters force the user back to the frameset, but my guess is the majority of frame sites don't do this, so google may have taken the overall opinion that it is better to downgrade these pages. In niche areas they will still rank well, but not against comparable non frame sites especially in more competitive sectors.
Frameset pages do naturally downgrade themselves. A site with an index link to a framset page then will dilute the pr to three or more frame pages. The real content (noframe content being dodgy) is now two links away from the index page. With a non frameset site, all the content is one click away and the pr flow is more focused and stronger.
If one of the frame pages for a frameset is the navigation, the spider see's a page with just links and no real content. This could have a disadvantage in 'taking the anchor text' seriously, having no onpage text support for the words.
Likewise, other frame pages may just be content, with no anchor text.... and we all know how google loves anchor text.
Enough said? Stick to static table pages :)
I would have to agree with MHes. Frames are bad news for SEO. It really screws up the PR passing and anchor text.
| This 43 message thread spans 2 pages: 43 (  2 ) > > |