Forum Moderators: open
My company is looking into getting a CMS and I want to know what to watch out for.
Thanks for your thoughts!
[example.com...]
It's the & that can cause pages not to be spidered, but even that is not a given. This url should not cause problems:
[example.com...]
With Apache's mod_rewrite it's possible to eliminate this problem altogether. Failing that you could look out for a CMS that creates simpler urls.
We have sites that have content management behind them - all of which we build from scratch, Completely customed to what the client needs.
I think it is more to do with what the CMS produces that could cause the problems rather than the actual CMS itself. As in the example above, if you have a page with a ridiculous amount of &something=something&something=something&something=something& etc, it might not get listed - but it would make no difference whether the URL was created manually or through a CMS. Its the &'s that apparently cause the problems.
[edited by: pageoneresults at 5:35 am (utc) on Nov. 24, 2004]
[edit reason] Fixed Side Scrolling [/edit]
CMSes are generally bad because they tend to overuse session and personalization information on the line. While it's true that most search engines handle dynamic URLs just fine, they do not handle session information properly.
Often, fixing it isn't as simple as just using mod_rewrite, because the CMS is dynamically creating the links as the content is created. You have to modify the code of the CMS itself.
CMS Matrix [cmsmatrix.org] is a good independent and objective resource for investigating your CMS options. Spiderability is one of the criteria they've reviewed.
Say for example, your site is about "game-x" and you use "cms y". Isn't it more likely that your site will be seen as a near duplicate to another site about "game-x" running on "cms y"? Particularly if the content on some of your pages is the same (in this case cheat codes, data feeds etc)?
It seems that if your site is already 20% similar (because of the interface/cms) to about 5000 other sites before you've even added a word of your own - aren't you starting at a disadvantage?
Just curious what the experts think.
because most of the time programmers design software and the majority of them know nothing about SEO, therefore they seldom take into consideration the implication of being spidered etc. Jake's quite right, that's a very good site.
>> While it's true that most search engines handle dynamic URLs just fine
be that as it may, would you ever do a site with nasty query strings Jake?
You're listenin to the search engineers too much ;)
would you ever do a site with nasty query strings Jake?
LOL, touche. Probably not. ;-)
Jake is right and wrong... it's good advice which may be ignored only if you're careful.
Many CMS systems use windows-based text editing panes (you know, the ones like you see in hotmail on IE, using execCommands).
The HTML created by those tools looks like the kind of garbage you get when you export a Word doc as HTML. You might as well just barf all over the screen.
CMS systems are not bad if they save time & streamline production, AND as long as they produce good clean source HTML&CSS. If you are so inclined, it's easy to make a database-driven CMS from scratch with SEO in mind.
I made one with PHP that can be configured to produce equisitely optimized results, but only if used with skill (the CMS does not produce the SEO, rather the SEO is in the way the CMS is used)
With a CMS you "can" lose control of on page aspects, Title, Meta Description and many of them without hacks have a session ID, long URLS and some have heavy HTML to Content Ratio.
What he is saying, is look around and compare the CMS's before your get stuck with a system that is not Search Engine Friendly and hurts your site rather than help. Even if it is easy to create content, if it can't be found what good is it?
Building your own CMS is a great deal of work but can be very beneficial if done well with SEO in mind.
If you are using a good CMS and have taken the time to set it up well, nobody will even that you are using one. You would be able to customize templates to pull generic or page specific meta tags, keyword stuffing etc... for any possible part of a web page.
they tend to overuse session and personalization information on the line
In my experience the best of today's CMSs do not have this problem.
I guess I'm not disagreeing with Jake, I just like to emphasize a different side of things.
using a CMS is bad news for SEO
Not so (as I think all of you are agreeing).
Provided the company/individual knows what seo factors need to be consiodered on a page, you simply ensure your developer builds you a bespoke system that does all of it for you - but leaves flexibility so that you can change things like page names, folder names, query strings to file names, suffixes from asp to .htm or .html etc etc etc
I know this works as I have over 100 sites using it in one form or another (all with unqiue content mind!)
The other classic we encounter is a client who has implemented a CMS to take over from a site which was partially content managed, because it 'makes things easier doing it in house', and then wants to completely rewrite it when they find they don't have the flexibility of working with a bespoke design(er) and programmer.
IMHO the main problem with CMS systems is that they are bought in under a misapprehension that all will be 'easy and quick' after this wonderous piece of software is up and running.
It is rarely 'easy' and seldom 'quick'...
The new one (Obtree, might only be known in Europe) produces great results and offers simple SEO features as well, like easy tag forms. What impresses me most are the statically looking URL's and the full control over path, length and structure of URI's, URL's.
Also, html code looks rather clean and lean.
A solid CMS vendor should feature such obvious concerns in their promotion and be able to clarify the situation with their potential customers.
Remember: education of content owners (publishers, editors) is key for SEO through, whenever you start thinking beyond technical barriers.
This is the route I chose.. maximum customizability -- You can make it do anything you want.
1. Most of the CMS generate content dynamicaly from their backend system , which leave little control of each and every page.
2. Most of them use template based page layout, which is generic in nature , customization for specific pages may be difficult.
3. As many people mentioned , they use URLs which is not spider friendly.
Remember: education of content owners (publishers, editors) is key for SEO through, whenever you start thinking beyond technical barriers.
I think the above point should be emphasized also. No good having good URLs if the content is not up to par.
Bad part is
------------
1) If you donot know how to write your meta tags in an efficient manner, then the navigation will come as SERP description ( home >> keypharse ) as this generally apprears towards top.
2) Content writers copy from MS Word or other text editors to the cms text editor which is normally WYSIWYG, the MS fonts are copied and the code becomes so Dirty and disorder, you need to teach the content writer a bit of <html>.
3) Personal touch of optimization is sometimes missing,
I have tried atleast 15 cms systems but none worked except the ones I started customizing. We have spend hours and hours making it SEO friendly (SE friendly URLs were just one part of it).
Last word of advice : every site should use cms for better ROI.
Aji
For those who are interested in creating top quality web sites - standards compliant, valid, rich semantic markup should be a requirement, not a "feature" for a CMS. Make sure the CMS outputs clean markup to help search engines index your site better, and also supports human-readable, memorable URLs. Don't ignore web standards - get the benefits of standards, such as XML, XSLT, CSS, XHTML, HTML, RSS/ATOM.
Sticky me and I'll try to point you in the right direction (no, I am not affiliated, just impressed ;-)
KOB
Let’s consider its essence: CMS = Content Management System
Does that implies auto-creating pages and Dynamic links, where is that defined, is it in writing?
Think about it from a programmer stand point of view
For example what does mine and how does it work (for all my sites, personal ones and clients’ one)
I first created a basic non-dynamic HTML site with pure, simple links and nav system
Then from that shell I transformed all the segments (I want to have added dynamism on) IN PHP OOP page and added my editable capability to those pages (texts and images)
Now I have a CMS that behaves fine within my bounds and criteria’s and no SEO links headache, it is still a CMS but no fuzz!
On what possible basis are you making such a sweeping statement?
It's entirely possible to use some sort of SSI-based system (Apache includes or slighly more complex house-style generators, which are not really CMS's in the true sense) to maintain sites using 'real' HTML files. I maintain at least 10,000 flat-file, extremely SEO-friendly, pages in under an hour a day using little more than that.
Using a CMS can take away the fine-tuning control that a highly professional site often needs, as well as merging two things that should be entirely separate: page generation and page management.
[example.com...]It's the & that can cause pages not to be spidered, but even that is not a given. This url should not cause problems:
[example.com...]
Even better...
There are ways to alleviate the whole "page.php" query altogether.
We've created a CMS that publishes database-managed content as an individual page based on keyphrase criteria defined for that page.
This may be helpful for those of you that can code ASP or PHP.
The administrative "add content" form has several of the basic fields for content data entry: Title, head, foot, article, link, etc. Then I have a keyphrase field which you would enter a hyphenated keyphrase.
Now you create a template for the articles. You can even have multiple templates which you can choose via a drop-down box in your CMS add content form.
In the template, you design <<<*A_Unique_Tag_Style*>>> to replace with the article ID, which is now defined -inside- the document as opposed to in a url querystring. (In ASP you would load the entire template inside the Replace() as a variable - don't know the equivalent in PHP.)
---
TEMPLATE EXAMPLE:
(/templates/template_content.asp)<%
'Open database connection
dim article_ID
article_ID="<<<*A_Unique_Tag_Style*>>>"
'Open content recordset using ID
%>
<html>
<body>
<p><%=RS("article_head")%></p>
<p><b><%=RS("article_title")%></b><BR>
<b><%=RS("article_link")%></b></p>
<p><%=RS("article")%></p>
<p><%=RS("article_foot")%></p>
<body>
</html>---
When you add your content to the database, there should be a simple administrative page which allows to to view, add, edit, delete, or PUBLISH each article on an individual or group basis. In ASP, this is accomplished using the File System Object (FSO).
Once you publish a page, it should load the template, replace the <<<*A_Unique_Tag_Style*>>> with the ID that was autonumbered inside the db table, and output it as a keyword-oriented page.
EXAMPLE:
---
OUTPUT EXAMPLE:
(keyword-phrase.asp)<%
'Open database connection
dim article_ID
article_ID="42"
'Open content recordset using ID
%>
<html>
<body>
<p><%=RS("article_head")%></p>
<p><b><%=RS("article_title")%></b><BR>
<b><%=RS("article_link")%></b></p>
<p><%=RS("article")%></p>
<p><%=RS("article_foot")%></p>
<body>
</html>---
Now, you're positioning "http://www.example.com/keyword-phrase.asp" instead of "http://www.example.com/page.php?var=keyword+phrase".
You can even go a bit further and use the keyword-phrase convention to pull up "keyword-phrase.gif", keyword-phrase.jpg" within the page layout, create a "print.asp?page=keyword-phrase.asp" type link, or whatever else you choose to enhance or promote the page.
I use the dynamic capability to detect spiders and then strip out extraneous content and instead serve an "abstract" of the content to the spider.
Isn't this considered a 'shady' practice? I seem to remember reading that G**gle specifically frowns upon delivering content to spiders that is different from the actual content on the user-viewed page...
www.example.com/product.php?section=1&23
or
www.exmaple.com/product/dvd/23
I would rather say the last one out loud over the phone!
Usability and accessibility issues might also be of some concern.
There are however, several companies that seem to be making the effort to address this issue in the last year, one in particular has been quite aggressive in designing its system with a heavy SEO/SEM slant.
I found this press release that I though might of interest to anyone who is dealing with these issues.
[ebizq.net...]
- Happy Thanksgiving!
URL issues are as simple as a an apache rewrite. You need to be able to control metatags, titles and headings at the page level, but that's just a couple more fields in the database and isn't any more difficult than the body text.In fact, I think a rewrite, once implemented can be better than hand coding page names (I can name each page whatever I like - optimized or not - because the mapping from the displayed URL to the actual page is just a database lookup).
HTML can be coded cleanly. And for the session info that the spiders dislike, all I do is check a list of spiders. If it matches, we don't serve any sort of cookie/session info. They get the same content (though clearly it's an immediate step to cloak from there).
The thing I like about CMS is that it seperates layout from content. I can do any layout SEO in one spot for everything, then manage my content seperately.
Yes, I'm top spot on some not horribly competitive phrases using a CMS.
