Welcome to WebmasterWorld Guest from 3.227.233.6

Forum Moderators: Robert Charlton & goodroi

Message Too Old, No Replies

Updating whole Website - Any precaution towards Google?

Before I am going to make a mistake, I'd rather have your opinion.

     
11:37 am on Apr 1, 2007 (gmt 0)

Junior Member

10+ Year Member

joined:Sept 28, 2002
posts:76
votes: 0


Currently I am completely revising my website. The website was made with fronrpage and recently I decided to update the whole website with a new look.

I changed the shared borders from FrontPage in dreamweaver to page top, side etc. Furthermore I removed the excessive amount of anchor text, H1 headings, FrontPage coding and more stuff that is not really appreciated anymore these days. I also adjusted for some pages the titles and descriptions in the Meta tags. The urls are all kept the same, so no loss of pages.

In general I would say I am heading in the right direction. However with Google you never know (I guess that is the Google Trauma in me). My question is, can I just remove all my old pages (400+) and replace them with my new pages? Or should I be careful to not get penalized by Google for some reason?

Thank you in advance for your input!

4:06 pm on Apr 1, 2007 (gmt 0)

Full Member

10+ Year Member

joined:Feb 1, 2006
posts:271
votes: 2


IMO you should take care with this

your site might be seen as a "new" website byG.
with all cnsequentises of it (sandbox)

why not try a few pages first?

4:21 pm on Apr 1, 2007 (gmt 0)

Preferred Member

10+ Year Member

joined:June 5, 2006
posts:352
votes: 0


No please don't do it (especially if you have new links) - I got seriously burnt doing it. The rule to follow is this: if you have traditionally published/updated x pages a week, just do 1.25X pages a week and keep doing it till you are done. Otherwise, Google does not like it and may put you in a "sandbox" - I was put in a 90 day sandbox after I completely redid the website with new link structure.
4:45 pm on Apr 2, 2007 (gmt 0)

Junior Member

10+ Year Member

joined:Sept 28, 2002
posts:76
votes: 0


Amazing to hear that. I find it weird to me that you have to be carefull with updating my website, since it could mean I could get sandboxed.

If you update a website or give it an overall updated look, you offer something better to the visitors. This is what bothers me with Google, that things that are a good thing, can result in bad results such as the ones you describe.

I now have a FP website, and some of the FP components will result in problems if I would upload some new pages.

5:00 pm on Apr 2, 2007 (gmt 0)

Preferred Member

10+ Year Member

joined:Oct 26, 2005
posts:426
votes: 1


I do not know if it would be a bad thing to update, but if you are not sure, just do some a/b tests on a few pages and see how it goes.
5:14 pm on Apr 2, 2007 (gmt 0)

Senior Member

WebmasterWorld Senior Member 10+ Year Member

joined:Jan 3, 2003
posts:785
votes: 0


Okay...so let's go over exactly what you are changing:

=> You are cleaning up the old messy Frontpage code - GOOD
=> You are making some visual changes on the site for layout / graphics / colors / images..?
=> You are changing out some page <titles> and meta tag information ... WHY?
=> You are retaining all your original (400) pages and their associated URLs (this implies that your internal link structure will remain intact along with any external links pointing into your site ... not to worry about this for your root URL...but if you have external links pointing to some of your internal pages...then you want to protect these for sure..) - GOOD
=> You Wrote:
Furthermore I removed the excessive amount of anchor text, H1 headings,...
When you are referring to anchor text ... do you mean links embedded in your body text ... or do you mean anchor text in your navigation...?
When you mention H1 headings...do you have only one per page ...?

5:23 pm on Apr 2, 2007 (gmt 0)

Junior Member

10+ Year Member

joined:Jan 16, 2004
posts:176
votes: 0


There should be no problem as long as you keep the same URLs, although of course changing the link structure can alter the way PR flows.

I recently redesigned a 600+ page site with no ill effects. I actually changed all the URLs and used an .htaccess file for 301 permanent redirects. Everything went smoothly. No change in ranking or traffic after three weeks.

6:40 pm on Apr 2, 2007 (gmt 0)

Junior Member

10+ Year Member

joined:Sept 28, 2002
posts:76
votes: 0


Thank you decaff, here are some answers to your questions:

You are making some visual changes on the site for layout / graphics / colors / images..?

Layout, graphics and colors: yes I change them all.

You are changing out some page <titles> and meta tag information ... WHY?

Well for some of them I noticed that I used a standard title, instead of a title that reflects the content of that page.

When you are referring to anchor text ... do you mean links embedded in your body text ... or do you mean anchor text in your navigation...?

Yes I removed lots of anchor text in my navigation,

When you mention H1 headings...do you have only one per page ...?

I removed most of the H1 headings, other then one in the top.

No urls of pages will change or deleted, these will all stay the same and keep the same content.

7:45 pm on Apr 2, 2007 (gmt 0)

Senior Member

WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month

joined:Feb 12, 2006
posts:2709
votes: 116


i redid my site a while ago and i'm doing better than ever now. i changed loads of things, though, so i can't say for sure which one helped me out the most, but here is a list of the most likely suspects:

1) reduced the amount of code, compared to text
- this was just a matter of getting rid of all the unnecessary divs, spans, <br>'s and inline CSS styles that you use when you're starting out.
this has a double benefit i reckon, because your actual text will gain more weight in the page algo, and more of your site will be spidered in one go, because it doesn't have to download as much (i gzipped all the pages as well, and introduced php caching)

2) removed every single piece of javascript off of the page, and stuck it in an external file.

3) used absolute positioning to move all of my actual text to the top of the HTML page, rather than having it come after all the navigation.

4) changed the main keyword so that it appears just once in the title, once in the description, once in the keywords, once in the <h1>, once in an image name, and once in the first paragraph... and just a very light smattering throughout the rest.

5) added a P3P policy, and ICRA label.
- these might not have had any effect, but with google banging on about "trust" all the time i figured that it couldn't hurt.

10:10 pm on Apr 2, 2007 (gmt 0)

Senior Member

WebmasterWorld Senior Member 10+ Year Member

joined:Nov 27, 2003
posts: 1648
votes: 2


If you are concerned about doing a slow changeover because you feel that the old frontpage will interact badly with the new pages, try it this way:
* do a wget of the entire site to get static html pages.
* turn off frontpage (whatever that takes, I don't know) completely on the site
* put up the static pages
* start updating the pages one by one with the new system

It will mean you can no longer do updates with frontpage, and the common-across-the-pages elements are now individual but this shouldn't be a problem because you are in the process of changing to a new system.

Hope that makes sense :)

11:13 pm on Apr 2, 2007 (gmt 0)

Full Member

10+ Year Member

joined:Feb 1, 2006
posts:271
votes: 2


what is a wget?
12:07 pm on Apr 3, 2007 (gmt 0)

Senior Member

WebmasterWorld Senior Member 10+ Year Member

joined:Nov 27, 2003
posts: 1648
votes: 2


wget - a tool to allow you to grab a copy of a website.
eg. It allows you to get each page of a generated site - if a site has 1000 pages all generated from a template and a database, wget will let you put the 1000 pages on your hard drive.
Very useful stuff :)
2:42 pm on Apr 3, 2007 (gmt 0)

Senior Member

WebmasterWorld Senior Member 10+ Year Member

joined:Apr 26, 2006
posts:1397
votes: 0


Currently I am completely revising my website...

Me, too. I'm also using .htaccess with redirects. No sign of trouble yet, but I'm being cautious. Is Google smart enough to read your .htaccess file to figure out "new" pages are old pages with new names? I find it annoying Google keeps the old Google cache for a renamed webpage, while it indexes new pages. So there's a ton of duplicate content. (Previously I've had old web pages, deleted months if not years ago, still cached and indexed. This just wastes the time of surfers, who keep getting 404 errors; I keeping seeing these in the error logs.) The Google algo should be smart enough not to issue a penalty for rapidly renamed pages with pretty much the same content. But I've been updating old content on the same pace as I'd been adding new pages.

p/g

12:24 am on Apr 5, 2007 (gmt 0)

Senior Member

WebmasterWorld Senior Member 10+ Year Member

joined:Nov 27, 2003
posts:1648
votes: 2


potentialgeek: google doesn't read your .htaccess file - its not publically viewable (on a correctly implemented web server)
Instead, the .htaccess file controls what happens to your page - think of it as a program
So, it will go like this:
* Googlebot requests page1.html
* Your server runs the .htaccess and responds '301 - see newpage.html' (normally the webserver responds '200 - here you go...')
* Googlebot thinks 'ah, 301 means a permanent move, I will do the standard processing I do with those'
* Presently (not straight away in googlebot's case) you will see Googlebot request newpage.html

Unfortunately Googlebot will also continue to request page1.html for a very long time.

hope it helps :)

7:09 am on Apr 6, 2007 (gmt 0)

Senior Member

WebmasterWorld Senior Member 10+ Year Member

joined:Apr 9, 2005
posts:1509
votes: 0


who keep getting 404 errors…

If your 301 redirect in your .htaccess is properly implemented, no request for the page from Google, or visitors should result in a 404. It should result in a 301.

You can only issue one status code per URI, unless you are using some sort of 'user-agent' switching, which would seem backward and purposeless in this case.

If you are running into an issue with 404 (Not Found) resources (which are not redirected to a 'new' location) being requested, you might implement a 410 (Gone) which indicates to a requesting agent the resource (document) should no longer be requested, because it has been 'permanently' or 'intentionally' removed.

A 404 only indicates the resource is 'Not Found' (at this time), but does not specify the resource has been 'permanently' or 'intentionally' removed, so the properly handling of a 404 is to re-request the resource periodically to see if it has become 200 'found'.

The following would indicate www.example.com/thepage.html has been intentionally removed and should not be re-requested:
RewriteRule ^thepage\.html$ - [G]

Justin