I’m very new to this and in the final stage of developing a mySQL based website. It will replace an existing ecommerce site that has all products programmed into the HTML. This site has always ranked well. Can anyone suggest a source of information to inform me how to ensure good ranking when all the products are in a database and not in the HTML?
The first thing is to make sure all of the pages of your site will have different page titles and meta description tags. This can be done by either using some of the database content in the tags, or creating fields to hold the information for each record.
You also should do keyword research, but that's another topic.
You should then use .htaccess or an ISAPI module to change the URLs of the site to remove the "?" and "&" and "=" from the urls, if you have them. These won't be a problem for Google and some other systems, but many smaller search engines can't or won't deal with them.
The other thing you will want to do is create some kind of redirection pages for most or all of your old URLs that will be deleted after you have the new site. This will prevent you from haveing what is known as "Traffic-Death Redesign".
Originally posted by Christian_SEO The first thing is to make sure all of the pages of your site will have different page titles and meta description tags. This can be done by either using some of the database content in the tags, or creating fields to hold the information for each record.
I’m not sure what you mean here. Since my pages are being generated by php code, they only exist in a generic form. They don’t really exist until someone clicks on a link.
You should then use .htaccess or an ISAPI module to change the URLs of the site to remove the "?" and "&" and "=" from the urls, if you have them. These won't be a problem for Google and some other systems, but many smaller search engines can't or won't deal with them.
That’s new to me. Thanks for the tip.
The other thing you will want to do is create some kind of redirection pages for most or all of your old URLs that will be deleted after you have the new site. This will prevent you from haveing what is known as "Traffic-Death Redesign".
I had forgotten about that. Thanks for reminding me.
I’m not sure what you mean here. Since my pages are being generated by php code, they only exist in a generic form. They don’t really exist until someone clicks on a link.
Right, but neither does your content...
Let's say you are selling fruit. You have one page showfruit.php that shows all the different kind of fruit you sell in the database.
When someone hits a URL for www.myfruit.com/showfruit.php?fruit=17, that offers your fine apples. The server pulls the page, adds in the product information and tells us about your fruit. You can also have it show us different optimization for each product.
Depending on the fields that you are using for the products, you can use some or all of this information to create page title tags and description tags that will be different for each type of fruit.
If the product description uses a field called fruit-type, then you would use this in the HEAD area of the page:
<.title>Our $fruit-type are always the best possible $fruit-type<./title>
I hope this helps, but post again if you have questions...
Originally posted by Christian_SEO
When someone hits a URL for www.myfruit.com/showfruit.php?fruit=17, that offers your fine apples. The server pulls the page, adds in the product information and tells us about your fruit. You can also have it show us different optimization for each product.
So, are you saying that the search engine crawler will actually travel through my website just like a user clicking on the links and then examine the generated pages? In that case I don’t really have a problem. I just need to take your advice about generating meaningful tags.
You should then use .htaccess or an ISAPI module to change the URLs of the site to remove the "?" and "&" and "=" from the urls, if you have them. These won't be a problem for Google and some other systems, but many smaller search engines can't or won't deal with them.
I highly recommend using some type of URL conversion to get rid of the "dynamic" characters as mentioned above. Depending on where your site gets traffic, this could restrict how much you get.
I had one client site, http://www.thejobspider.com where the client is using the ISAPI URL conversion on my recommendation. He did not convert the "=" at first. After seeing some problems a few months later he went back in and changed that also and noticed quite a difference in traffic and how much of his site gets spidered. For most sites it's well-worth the effort, in my opinion...
Originally posted by Christian_SEO Well, yes and no...
I highly recommend using some type of URL conversion to get rid of the "dynamic" characters as mentioned above. Depending on where your site gets traffic, this could restrict how much you get.
Ok, so I have been investigating ISAPI URL conversion. The information I have read sounds like I’m supposed to install something on the server. I’m paying a web hosting service to host my website. It’s a Linux/Apache/PHP system. Often the references sound like they refer to Windows based systems. (BTW, I’m developing on a Mac) Also, the information I have found assumes an a prioiri (haven’t used that expression since my Engineering days) level of knowledge that I do not possess. Can you suggest a source of beginner level information? I understand the need to replace my links which look like
Wedding_Dept/category.php?par_cat_id=2
with something that looks like
Wedding_Dept/category/par_cat_id/2
What I don’t understand is who or what is converting this back so my php code will understand the link?
By the way, thanks for the help so far. The information has been very useful.
On a Windows server you do have to install something and if you are hosted somewhere they may not what to make the change to the server (althought it's a huge benefit for you, the customer).
On a UNIX system you have to edit something called a .htaccess file.
In this configuration file you can specify the "rules" on how the conversions are made. The example you gave was good, but I would avoid "_" in favor of "-". "/", or "~" since they "break up" the words that are use. If you use a "_" it is seen as part of a "word". (two-words is seen as two words, but two_words is seen as one and will not match searchs as well.
I'm NOT a UNIX/Linux guy yet, but if you do some searches on the use of .htaccess and URL conversion or URL replacement, you should find what you need. If not, let me know and I'll dig something up for you... :-)
Thanks guys for your help. I have contacted my web host and they said the following:
====
We run apache but not apache for windows which is a prerequisite for the isapi module. I think what your looking for in the linux environment is apache with mod_rewrite. We actually do not currently support that however we will be rolling out an upgrade including the mod_rewrite module to all of our servers between march 23 and march 25th so if you can wait untill then mod_rewrite should work for you.
====
I did a quick search on mod_rewrite and it seems to relate to .htaccess. I need to study this and see if I can figure out how to use this.
which generates a link with a ? and an = in it. So, I would change the code and replace those characters with something else. Then when the link is clicked on and the code arrives at the web hosting server, the URL would be translated back to use the ? and = and the URL would be delivered to my php code.
Ooops! I guess it was mod-rewrite and not .htaccess... well, I said I'm not a UNIX guy and that proves it...!
No, with these programs you do NOT have to change your code.
Here's a simple way to explain what happens:
When the server sends a page, it looks at the URLs on the page.
When it sees "www.domain.com/product.php?code=11"
It says, oops, there's work to do, and turns it into something like this:
"www.domain.com/product.php/code~11"
When someone clicks on the link and requests the page from the server, the process is reversed.
I think you CAN change your site URLs if you wanted to, so they were "right" from the start, but I don't think you gain much and if the conversion process changes then your code would have to change. The nice thing about this is that if the server is sent the old or new URL format, they both work, so whatever links you have will not be lost and disabled by this change.
Originally posted by Christian_SEO When the server sends a page, it looks at the URLs on the page.
When it sees "www.domain.com/product.php?code=11" It says, oops, there's work to do, and turns it into something like this:
"www.domain.com/product.php/code~11"
When someone clicks on the link and requests the page from the server, the process is reversed.
That’s great! I guess I was assuming it would only notice the URL when it was being delivered to the server from the browser, but not actually look for URL’s in the page on the way to the browser. Now I just have to wait for .htacess to be available at my host and I'm set to go.
Thanks to both of you for your very helpful responses.
Reply With Quote
Bookmarks