Improving your search-engine status - Part 3
There are some ways of organising your web pages, and your web site, that can cause problems for your search-engine ratings. So let's look at them, and what to do about them (note - this section assumes a little more technical knowledge than previous sections).
Most search-engine aren't very good at navigating through pages which are made up of frames. They tend to get stuck on 'frameset definition' pages, which just describe the size and source of each frame in a given page. So the content and meta information in these frames may not be gathered by the search-engine crawler.
The general solution to this problem, if you can't do without frames, is to provide, for each frameset definition, a 'noframes' page containing the appropriate material for the search engine. There are some issues that have to be addressed, however, in order to set this solution up properly. The following article provides some useful insight:
http://searchenginewatch.com/webmasters/frames.html
We noted in the previous section that the page content at the top of the page may be taken by a search-engine crawler to be more significant than that further down. So if you clutter up the tops of your pages with excess material, it may affect your search rating for the searches you care about.
A common example of this issue is the placement of JavaScript code. For understandable reasons, scripts are often placed high in pages, where they can distract a crawler from the text you want to be noticed. It is generally possible, however, to relocate scripts within a page, or to import the code using a link.
Another issue relates to tables. Suppose, for example, that - having eschewed frames - you lay out your page using a table, with the navigation material on the left. In the usual way of specifying tables, the navigation material will be specified in the code above the general contents, and will thus be taken as more important.
There are some technical tricks to get around this problem, however. One of these can be found at:
http://www.submitcorner.com/Guide/Improve/optimizemain.shtml
In some sites, most of the content is generated from a database in response to queries. Examples would include sites which allow you to search for local IT firms from data held in storage. It may seem obvious, but it is worth taking note of the fact that none of this information will be picked up by a search-engine crawler.
What, though, if you do want your information picked up by a search-engine crawler, but the information powering your site periodically changes? Here it may be impossible to rewrite all the pages on a regular basis.
One answer would be to write a server-side program which queries your database and outputs static HTML pages from it. This program could be run every time the database changes, or be set to run every day at 3am, etc.
Search-engine crawlers tend to be leery of pages which automatically redirect users elsewhere, and may refuse to index or follow them. So, for instance, suppose that you have moved your site, and set up an automatic redirect from the address of the old site. You may have to resubmit your site at the new address, since crawlers will not follow the redirection.
The reason that crawlers don't like redirects is quite interesting. One of the black arts developed by webmasters has been to develop 'bridge pages' for submission to search engines. The idea behind such pages is this. It is possible, based upon intimate knowledge of a search-engine's indexing algorithm, to design pages which index extremely well and thus come out high on the search results list. But such pages tend not to be user-friendly. So, those who submit them also ensure that people who navigate to them through search results are swiftly redirected to the user-friendly pages.
Search-engines don't approve of this process, however, since the searcher could be taken to any page, regardless of its contents. Hence search-engine crawlers may refuse to index these pages, and may balk at redirection generally.
next part
|
