The "Freeola Customer Forum" forum, which includes Retro Game Reviews, has been archived and is now read-only. You cannot post here or create a new thread or review on this forum.
Having missed all the excitment over the last few days, I return with a monstor post. This is largely my own work, but I have drawn upon many resources including the excellent http://www.searchenginewatch.com.
This is my first Masterclass, a thread where I write about something I think I know about, and where you guys get the chance to pick my brains if you like. This thread (and others to come if you want) serves 2 purposes to my mind, I share my knowledge with you, but you share yours with me too, that way, we all know a little bit more about the crazy web.
My first masterclass is on search engines, this is a report I first started writing about 3 years ago, but which has changed a few times since, and may be slightly out of date, but hey, you can't win 'em all.
Class begins.... this article is Copyright to me. It has been distributed to a few companies but contains nothing you cannot research for yourself on the Internet. If you blatently copy and paste it across the internet and fob it off as your own work you may find yourself in some hot water.
-----------------------------------------------------------
Copyright 2002 - All Rights Reserved
Contact [email protected] for permission to use this text, or portions of it anywhere else.
What is a Search Engine?
In the early days, when the web was just beginning, it was not hard to find a site that had what you wanted, as there were not many sites to look at in the first place. As the Internet (and therefore the web) grew, domain names were snapped up, and web sites became more and more commonplace. Now there was a plethora of information at the surfers' fingertips, the problem was not "is there a web site about widgets?" but was "where can I find a web site about widgets". Enter the Search Engines.
Essentially, a search engine is a database of web sites. Search engines allow users to interrogate the database for specific, or even non-specific information, the database should return a list of web sites based on the users' search criteria, the user can then select the web site they want and go merrily surfing off.
There are several different search engines, some of which are simply providing a different front-end (or interface) to a single back-end database. There are some search engines which do maintain a separate database of web sites but which may rely upon other services for data if their own is not sufficient to provide results to a search.
* Search engine - a big database which holds information on web sites, users can query the database for sites relevant to their needs. (http://www.northernlight.com is an example of a search engine)
* Search directory - like a search engine, the difference being that the directory is designed in such a way that users are supposed to "drill down" the directory structure (i.e., entertainment / tv / bbc1 / Monday / etc...) to reach the specific information that they want. (http://www.yahoo.com is the most famous search directory, but which is often mistaken for a search engine).
* Meta-search engine - type a query into a meta-search engine and it will send your request to multiple other search engines, the results of which will then be collated and displayed. A meta-search engine does not have its' own database, but relies upon others. (http://www.ask.co.uk for a good meta-search engine)
All of the above will be termed search engines from now on
Very few search engines exist to solely provide users with a search service. Many search engines exist to make money from banner adverts or other "services" or "buying opportunities" arranged around search results in such a way to distract the user from the results. It is a fair guess to assume that at least as much time money and effort is spent trying to distract users to banner ads, etc, as is spent searching out, collecting and categorizing web sites to put in the database. Google.com and dmoz.org are fairly unique search engines in that they exist simply to serve other search engines and do not have paid-for advertising.
How do search engines work?
As was mentioned before, there are many search engines out there and therefore many different methodologies are employed to:-
1. Find a site to register
2. Look at and "evaluate" the sites' worthiness to be included into the search engines' database.
3. Allow the user to search.
4. Display the site to the user after a search has pulled a specific web site from its' database.
1: Find a site to register
This first step is easy. Either a search engine uses a registration form for users to fill in and submit the site for evaluation, or the search engine employs "spiders". Spiders can trawl the web looking for sites independently of whether they are submitted for evaluation or not. Some search engines may combine both the form and spider technique, simply queuing up potential web sites that have been submitted via a form, until the spider gets around to looking at them ("spidering").
2: Evaluation of the web site
The second step is perhaps the most controversial. Once a site has been spidered, it must then be decided what category the site goes in (if this has not already been defined by a human submission form), to what extent Exactly how do search engines deem if a site is worthy or not to be indexed? There are many means at a search engines' disposal to decide upon this:-
* Meta-tags - there are 2 main types of meta-tags,
1. Keywords, in which you place the words that are most relevant to your site.
2. Description, which is usually a meaningful sentence about the web site. These are not visible to the average web surfer without looking at the source code to the web page.
* Document Title - every HTML document on your web site should have a title. This title is displayed in the blue bar across the top of a users' web browser.
* Body text - the visible text on the web page.
* Alt tags - ever held your mouse over a button or image for a few seconds, to see a little box pop-up with a description. Alt-tags are within the HTML source code of a web page but which define what the little pop-up boxes contain.
* Links to and from your site - spiders can follow the links that are defined within HTML to jump to other web pages and web sites altogether. Some spiders check to see if you link to sites already contained within their database, if you do, what category of site do you link to, what is the standing / popularity / quality of that site within the database. Similarly, a web site may be judged upon the web sites which link to it. If popular / high ranking sites link to a specific web site, and vice versa, the web site in question may be given a better position based upon this.
* Human evaluation - Sites like Yahoo! and http://dmoz.org/ rely upon humans (or "editors") to evaluate a site. If a submission is correctly filled out and the site does what it purports to do then (hopefully) the human editor will allow the site into the index.
* Domain name - what is your domain name? Does it match the theme of the site?
* Other points -
1. It has been suggested that some search engines actively use software to scan Internet newsgroups for relevant links to web sites
2. Once a spider (or editor) has visited a site, and the site matches the set criteria, this still does not mean that the site is placed in the search engine index immediately, it may be between 24 hours to six months before the site appears in the index at all, never mind near the "top" of the listing.
3. Some spiders revisit sites to ensure that they still exist, in doing so they may check the time at which the site was last updated. If the site has not been actively updated for an extended period of time it may be assumed that the data held on the site is less relevant or up-to-date than a freshly submitted site. The older site may be penalized and given a lower position in the index.
4. Many search engines like to publish that they have 20 million pages indexed and at the disposal of searchers, however quantity does not mean quality. How many times have you searched for "greenhouse" only to see "XXX rated" type websites appear in the results? (Yes, this has happened!)
5. Search engines may be using software to detect which web sites are being clicked onto the most from their listings. Those sites that are constantly avoided by users may be penalized.
6. The order in which keywords appear in meta-tags, title and body text has a bearing upon the weighting given to them, ie, words which appear first are deemed more important than words which appear after
7. The size of words in the body text has a bearing, large titles at the beginning of a document are given more precedence, and are sometimes used in the listings that search engines give.
With the exception of dmoz and yahoo!, it is estimated that differing search engines rely upon different combinations of the above to rank a web site. For instance, one search engine may simply take your meta-tags and title and from those decide upon where you should be placed, another search engine may place emphasis upon your meta-tags and body text, yet another may look at the amount of body text you have and the words contained within it, but will ignore the title to your document.
There is a phenomenal amount of calculations (or algorithms) which search engines can apply to web pages to work out if it gets into the index, and under what category. Educated guesses can be made as to which techniques a spider uses to index a web site, but unless search engines make public the exact way in which they decide how to rank a web site, it will be impossible to tell how it is done with a 100% degree of certainty.
In the early days, webmasters worked out that top rankings could be achieved by using "spamming" techniques such as using characters such as !, @, and others to fool the search engine into giving the site a higher alphabetical order, or hiding white coloured text on white background that contained multiple keywords to fool the search engine that enormous emphasis was being placed upon this subject. Whilst this has worked in the past, the top search engines have added anti-spamming techniques to their spiders, to the point where it is now not worth trying to spam a search engine as it will probably be detrimental to any listing, or even future attempt at a listing that a web site may have.
3: Allowing the user to search
Even if the first two stages are completed successfully, there is still the matter of web surfers being able to pinpoint a single web site amongst the many millions that can be held in the search engines index. Many users simply enter a few keywords, hit the search button and expect results, which is the simplest way of going about a search. This is akin to going into a library and saying to the librarian 'cooking' or 'gardening'. A more sophisticated searcher will start to use advanced Boolean logic functions available with some search engines that help to distill the database results even further. Instead of asking for "cooking", the user might ask for sites pertaining to "cooking" AND "Chinese", or they may use speech marks to denote that they want to view a site that contains information on "18th century Chinese cooking". Books have been written on the best ways of getting results from search engines, internet magazines are running tutorials on the how to use search engines effectively, so generally, public awareness about using advanced search functions is being raised.
4: Displaying Search Results
This final part of the effort is the part that should bear the fruits of fairly intensive labour. A common misconception about the results that search engines throw up is that there is a single list of web sites about a certain subject, and your specific site is either at the "top" or the "bottom". Search engines may display results in a manner of different ways:-
1. Alphabetically - A to Z, probably based on a sites' title or meta description.
2. Percentage based rating - sites that are deemed to closely match your search are given a higher percentage rating. Search engines like this are highly dynamic, applying algorithms to web pages in the database on-the-fly to come up with something (based upon meta-tags, title, body text etc) that it assumes is the best match to your search. Using this method, the same site may achieve wildly differing results for specific keywords. A site about called "cooking4u.com" may achieve really good results for "recipes" but not for the actual word "cooking". Both are still valid words used when searching for new dishes to eat. Again, the same site may achieve different results on a different search engines for the same keyword depending upon the internal calculations applied to the web site at the time of search.
3. Other internal search criteria - sites are not alphabetically listed, nor are they given a percentage mark, users just trust that the sites returned are relevant to the search conducted. It is likely that sites like these have their own version of the percentage based system as outlined above, but just do not show the final percentage that is given to the site.
The frustrating part of this is knowing that your web site has been spidered, and is contained within the index, but is simply not appearing for the keywords that you feel it should.
Best Practice
Given that we are aware of all the issues, how can we best promote a web site to the search engines? The first step is realizing that a "top" rating for specific keywords will not happen overnight, it will take a sustained assault upon the search engines to achieve anything like good results.
Upon first submission, sites may be indexed by search engines, but will almost certainly not perform too well except in a minority of cases. Perhaps after 6-12 months of constant re-submission and marketing the site will begin to achieve something like a high presence in the search engines. Think of it as market share. If you start a new company (or web site) up, you will have no market share, you must build market share up through hard work, and you cannot expect success to happen immediately upon start-up. So it is with web sites and search engines. It is also important to understand that achieving a good rating in search engines will not make or break a web site, search engines are but one piece of the web marketing jigsaw.
With this in mind, here are some search engine ratings tips:-
* Start thinking keywords at the design stage. Search engine marketing is not something to think about right at the end of the web site design and production cycle. Keywords should be defined and sprinkled about the web site pages during production.
* Only submit the site when it is fully operational, "under construction" signs are extremely undesirable.
* Is it possible to split the site into different sections, or sub-sites, (for instance, different products) and promote each section of the site individually to the search engines? Use different meta-tags and keywords for each unique section.
* Don't overdo the marketing slang. If you make caravans, use "caravans" as your keyword, not "holiday homes".
* Try not to get too technical - use words that the majority of people will know and use and understand.
* Get technical! Contrary to the above, if you are selling a highly specific product, to a specific group of people, ensure that you use the technical words that those people would use to find your product.
* Avoid ego trips - the Internet is famous for allowing local business reach a global audience, but the global audience probably does not know your business name. People tend too search using words like "playstation" or "gardening" or "caravan", they may not be looking for a specific company, but rather a specific product, so if you concentrate on pushing your corporate name onto the search engines, you are missing out on millions of people who do not know your company name, but who are looking to buy a product which you have.
* If possible, buy your company domain name as a .com and .co.uk domain. Many search engines give a higher priority to .com domain names (probably American bias, but also they expect .coms to be bigger and better then local regional domains.)
* If you market a range of products with trademark or individual names, try to buy the domain names of the products and point those domain names towards your main site (or sub-sites). This serves three distinct purposes.
1. Protect your trademark or product / intellectual copyright
2. The search engines will treat each new domain name as a separate website, so you stand a greater chance of appearing in the index through normal spidering activities. Your main domain name may not even fair as well as the other "satellite" domains for certain keywords, but does it matter as long as the site as a whole is doing well?
3. Many people randomly type in domain names into their browser, guessing that what they are looking for has a website under a specific domain. The chances of a site being found in this way are greatly increased if there are a number of relevant domain names pointing to it.
If you can purchase domain names that contain some of your keywords, this too can help to gain a better stronghold in the search engine indexes.
* Re-submit - about once a month make a habit of re-submitting the site to search engines. Contrary to what many search engines would like you to believe, there is a limit to how many web pages they can store in a database, eventually some web sites are removed from the index to make way for new listings. Constant re-submission means that every time the database is updated, you can be fairly sure that even if some of your pages are kicked out of the index, you have re-submitted the site so the spider should have allowed some more of your pages back into the index.
* Update the web site often - as mentioned before, spiders re-visit web sites to make sure they still exist, but also to ensure that updates are taking place. If a spider detects that a web site is getting dated it may decide to drop it from the index altogether. Sites should be updated regardless from a purely marketing point of view. Nobody wants to keep returning to a site with the same old information week after week. The truly great web sites are those which are never completed, but which are fluid and dynamic, and keep evolving week after week.
* Don't use frames, as spiders have difficulty with them. Those spiders that can cope with frames will probably index pages contained within the frames, but which will not make sense to a browser without the other frames around it to complete the picture.
* If the look and feel of the site is more important then anything else, use "bridge" pages instead of submitting your home page.
-------------------------------------
Copyright 2002 - All Rights Reserved
Contact [email protected] for permission to use this text, or portions of it anywhere else.
This is my first Masterclass, a thread where I write about something I think I know about, and where you guys get the chance to pick my brains if you like. This thread (and others to come if you want) serves 2 purposes to my mind, I share my knowledge with you, but you share yours with me too, that way, we all know a little bit more about the crazy web.
My first masterclass is on search engines, this is a report I first started writing about 3 years ago, but which has changed a few times since, and may be slightly out of date, but hey, you can't win 'em all.
Class begins.... this article is Copyright to me. It has been distributed to a few companies but contains nothing you cannot research for yourself on the Internet. If you blatently copy and paste it across the internet and fob it off as your own work you may find yourself in some hot water.
-----------------------------------------------------------
Copyright 2002 - All Rights Reserved
Contact [email protected] for permission to use this text, or portions of it anywhere else.
What is a Search Engine?
In the early days, when the web was just beginning, it was not hard to find a site that had what you wanted, as there were not many sites to look at in the first place. As the Internet (and therefore the web) grew, domain names were snapped up, and web sites became more and more commonplace. Now there was a plethora of information at the surfers' fingertips, the problem was not "is there a web site about widgets?" but was "where can I find a web site about widgets". Enter the Search Engines.
Essentially, a search engine is a database of web sites. Search engines allow users to interrogate the database for specific, or even non-specific information, the database should return a list of web sites based on the users' search criteria, the user can then select the web site they want and go merrily surfing off.
There are several different search engines, some of which are simply providing a different front-end (or interface) to a single back-end database. There are some search engines which do maintain a separate database of web sites but which may rely upon other services for data if their own is not sufficient to provide results to a search.
* Search engine - a big database which holds information on web sites, users can query the database for sites relevant to their needs. (http://www.northernlight.com is an example of a search engine)
* Search directory - like a search engine, the difference being that the directory is designed in such a way that users are supposed to "drill down" the directory structure (i.e., entertainment / tv / bbc1 / Monday / etc...) to reach the specific information that they want. (http://www.yahoo.com is the most famous search directory, but which is often mistaken for a search engine).
* Meta-search engine - type a query into a meta-search engine and it will send your request to multiple other search engines, the results of which will then be collated and displayed. A meta-search engine does not have its' own database, but relies upon others. (http://www.ask.co.uk for a good meta-search engine)
All of the above will be termed search engines from now on
Very few search engines exist to solely provide users with a search service. Many search engines exist to make money from banner adverts or other "services" or "buying opportunities" arranged around search results in such a way to distract the user from the results. It is a fair guess to assume that at least as much time money and effort is spent trying to distract users to banner ads, etc, as is spent searching out, collecting and categorizing web sites to put in the database. Google.com and dmoz.org are fairly unique search engines in that they exist simply to serve other search engines and do not have paid-for advertising.
How do search engines work?
As was mentioned before, there are many search engines out there and therefore many different methodologies are employed to:-
1. Find a site to register
2. Look at and "evaluate" the sites' worthiness to be included into the search engines' database.
3. Allow the user to search.
4. Display the site to the user after a search has pulled a specific web site from its' database.
1: Find a site to register
This first step is easy. Either a search engine uses a registration form for users to fill in and submit the site for evaluation, or the search engine employs "spiders". Spiders can trawl the web looking for sites independently of whether they are submitted for evaluation or not. Some search engines may combine both the form and spider technique, simply queuing up potential web sites that have been submitted via a form, until the spider gets around to looking at them ("spidering").
2: Evaluation of the web site
The second step is perhaps the most controversial. Once a site has been spidered, it must then be decided what category the site goes in (if this has not already been defined by a human submission form), to what extent Exactly how do search engines deem if a site is worthy or not to be indexed? There are many means at a search engines' disposal to decide upon this:-
* Meta-tags - there are 2 main types of meta-tags,
1. Keywords, in which you place the words that are most relevant to your site.
2. Description, which is usually a meaningful sentence about the web site. These are not visible to the average web surfer without looking at the source code to the web page.
* Document Title - every HTML document on your web site should have a title. This title is displayed in the blue bar across the top of a users' web browser.
* Body text - the visible text on the web page.
* Alt tags - ever held your mouse over a button or image for a few seconds, to see a little box pop-up with a description. Alt-tags are within the HTML source code of a web page but which define what the little pop-up boxes contain.
* Links to and from your site - spiders can follow the links that are defined within HTML to jump to other web pages and web sites altogether. Some spiders check to see if you link to sites already contained within their database, if you do, what category of site do you link to, what is the standing / popularity / quality of that site within the database. Similarly, a web site may be judged upon the web sites which link to it. If popular / high ranking sites link to a specific web site, and vice versa, the web site in question may be given a better position based upon this.
* Human evaluation - Sites like Yahoo! and http://dmoz.org/ rely upon humans (or "editors") to evaluate a site. If a submission is correctly filled out and the site does what it purports to do then (hopefully) the human editor will allow the site into the index.
* Domain name - what is your domain name? Does it match the theme of the site?
* Other points -
1. It has been suggested that some search engines actively use software to scan Internet newsgroups for relevant links to web sites
2. Once a spider (or editor) has visited a site, and the site matches the set criteria, this still does not mean that the site is placed in the search engine index immediately, it may be between 24 hours to six months before the site appears in the index at all, never mind near the "top" of the listing.
3. Some spiders revisit sites to ensure that they still exist, in doing so they may check the time at which the site was last updated. If the site has not been actively updated for an extended period of time it may be assumed that the data held on the site is less relevant or up-to-date than a freshly submitted site. The older site may be penalized and given a lower position in the index.
4. Many search engines like to publish that they have 20 million pages indexed and at the disposal of searchers, however quantity does not mean quality. How many times have you searched for "greenhouse" only to see "XXX rated" type websites appear in the results? (Yes, this has happened!)
5. Search engines may be using software to detect which web sites are being clicked onto the most from their listings. Those sites that are constantly avoided by users may be penalized.
6. The order in which keywords appear in meta-tags, title and body text has a bearing upon the weighting given to them, ie, words which appear first are deemed more important than words which appear after
7. The size of words in the body text has a bearing, large titles at the beginning of a document are given more precedence, and are sometimes used in the listings that search engines give.
With the exception of dmoz and yahoo!, it is estimated that differing search engines rely upon different combinations of the above to rank a web site. For instance, one search engine may simply take your meta-tags and title and from those decide upon where you should be placed, another search engine may place emphasis upon your meta-tags and body text, yet another may look at the amount of body text you have and the words contained within it, but will ignore the title to your document.
There is a phenomenal amount of calculations (or algorithms) which search engines can apply to web pages to work out if it gets into the index, and under what category. Educated guesses can be made as to which techniques a spider uses to index a web site, but unless search engines make public the exact way in which they decide how to rank a web site, it will be impossible to tell how it is done with a 100% degree of certainty.
In the early days, webmasters worked out that top rankings could be achieved by using "spamming" techniques such as using characters such as !, @, and others to fool the search engine into giving the site a higher alphabetical order, or hiding white coloured text on white background that contained multiple keywords to fool the search engine that enormous emphasis was being placed upon this subject. Whilst this has worked in the past, the top search engines have added anti-spamming techniques to their spiders, to the point where it is now not worth trying to spam a search engine as it will probably be detrimental to any listing, or even future attempt at a listing that a web site may have.
3: Allowing the user to search
Even if the first two stages are completed successfully, there is still the matter of web surfers being able to pinpoint a single web site amongst the many millions that can be held in the search engines index. Many users simply enter a few keywords, hit the search button and expect results, which is the simplest way of going about a search. This is akin to going into a library and saying to the librarian 'cooking' or 'gardening'. A more sophisticated searcher will start to use advanced Boolean logic functions available with some search engines that help to distill the database results even further. Instead of asking for "cooking", the user might ask for sites pertaining to "cooking" AND "Chinese", or they may use speech marks to denote that they want to view a site that contains information on "18th century Chinese cooking". Books have been written on the best ways of getting results from search engines, internet magazines are running tutorials on the how to use search engines effectively, so generally, public awareness about using advanced search functions is being raised.
4: Displaying Search Results
This final part of the effort is the part that should bear the fruits of fairly intensive labour. A common misconception about the results that search engines throw up is that there is a single list of web sites about a certain subject, and your specific site is either at the "top" or the "bottom". Search engines may display results in a manner of different ways:-
1. Alphabetically - A to Z, probably based on a sites' title or meta description.
2. Percentage based rating - sites that are deemed to closely match your search are given a higher percentage rating. Search engines like this are highly dynamic, applying algorithms to web pages in the database on-the-fly to come up with something (based upon meta-tags, title, body text etc) that it assumes is the best match to your search. Using this method, the same site may achieve wildly differing results for specific keywords. A site about called "cooking4u.com" may achieve really good results for "recipes" but not for the actual word "cooking". Both are still valid words used when searching for new dishes to eat. Again, the same site may achieve different results on a different search engines for the same keyword depending upon the internal calculations applied to the web site at the time of search.
3. Other internal search criteria - sites are not alphabetically listed, nor are they given a percentage mark, users just trust that the sites returned are relevant to the search conducted. It is likely that sites like these have their own version of the percentage based system as outlined above, but just do not show the final percentage that is given to the site.
The frustrating part of this is knowing that your web site has been spidered, and is contained within the index, but is simply not appearing for the keywords that you feel it should.
Best Practice
Given that we are aware of all the issues, how can we best promote a web site to the search engines? The first step is realizing that a "top" rating for specific keywords will not happen overnight, it will take a sustained assault upon the search engines to achieve anything like good results.
Upon first submission, sites may be indexed by search engines, but will almost certainly not perform too well except in a minority of cases. Perhaps after 6-12 months of constant re-submission and marketing the site will begin to achieve something like a high presence in the search engines. Think of it as market share. If you start a new company (or web site) up, you will have no market share, you must build market share up through hard work, and you cannot expect success to happen immediately upon start-up. So it is with web sites and search engines. It is also important to understand that achieving a good rating in search engines will not make or break a web site, search engines are but one piece of the web marketing jigsaw.
With this in mind, here are some search engine ratings tips:-
* Start thinking keywords at the design stage. Search engine marketing is not something to think about right at the end of the web site design and production cycle. Keywords should be defined and sprinkled about the web site pages during production.
* Only submit the site when it is fully operational, "under construction" signs are extremely undesirable.
* Is it possible to split the site into different sections, or sub-sites, (for instance, different products) and promote each section of the site individually to the search engines? Use different meta-tags and keywords for each unique section.
* Don't overdo the marketing slang. If you make caravans, use "caravans" as your keyword, not "holiday homes".
* Try not to get too technical - use words that the majority of people will know and use and understand.
* Get technical! Contrary to the above, if you are selling a highly specific product, to a specific group of people, ensure that you use the technical words that those people would use to find your product.
* Avoid ego trips - the Internet is famous for allowing local business reach a global audience, but the global audience probably does not know your business name. People tend too search using words like "playstation" or "gardening" or "caravan", they may not be looking for a specific company, but rather a specific product, so if you concentrate on pushing your corporate name onto the search engines, you are missing out on millions of people who do not know your company name, but who are looking to buy a product which you have.
* If possible, buy your company domain name as a .com and .co.uk domain. Many search engines give a higher priority to .com domain names (probably American bias, but also they expect .coms to be bigger and better then local regional domains.)
* If you market a range of products with trademark or individual names, try to buy the domain names of the products and point those domain names towards your main site (or sub-sites). This serves three distinct purposes.
1. Protect your trademark or product / intellectual copyright
2. The search engines will treat each new domain name as a separate website, so you stand a greater chance of appearing in the index through normal spidering activities. Your main domain name may not even fair as well as the other "satellite" domains for certain keywords, but does it matter as long as the site as a whole is doing well?
3. Many people randomly type in domain names into their browser, guessing that what they are looking for has a website under a specific domain. The chances of a site being found in this way are greatly increased if there are a number of relevant domain names pointing to it.
If you can purchase domain names that contain some of your keywords, this too can help to gain a better stronghold in the search engine indexes.
* Re-submit - about once a month make a habit of re-submitting the site to search engines. Contrary to what many search engines would like you to believe, there is a limit to how many web pages they can store in a database, eventually some web sites are removed from the index to make way for new listings. Constant re-submission means that every time the database is updated, you can be fairly sure that even if some of your pages are kicked out of the index, you have re-submitted the site so the spider should have allowed some more of your pages back into the index.
* Update the web site often - as mentioned before, spiders re-visit web sites to make sure they still exist, but also to ensure that updates are taking place. If a spider detects that a web site is getting dated it may decide to drop it from the index altogether. Sites should be updated regardless from a purely marketing point of view. Nobody wants to keep returning to a site with the same old information week after week. The truly great web sites are those which are never completed, but which are fluid and dynamic, and keep evolving week after week.
* Don't use frames, as spiders have difficulty with them. Those spiders that can cope with frames will probably index pages contained within the frames, but which will not make sense to a browser without the other frames around it to complete the picture.
* If the look and feel of the site is more important then anything else, use "bridge" pages instead of submitting your home page.
-------------------------------------
Copyright 2002 - All Rights Reserved
Contact [email protected] for permission to use this text, or portions of it anywhere else.
Mr. Nice Guy wrote:
> Ummmmm thanks for that. I too, might take everything from my site one day and
> post it on the forums, just in a bid to get my word count up.......
You might notice that my own site has nothing to do with web design, this is something that I researched and wrote a while ago, and just thought I would share what I thought with the forum.
My word count is already high enough, what actually I need is to get my post count up, that way I can be a regular and talk down to the newbies too.
> Ummmmm thanks for that. I too, might take everything from my site one day and
> post it on the forums, just in a bid to get my word count up.......
You might notice that my own site has nothing to do with web design, this is something that I researched and wrote a while ago, and just thought I would share what I thought with the forum.
My word count is already high enough, what actually I need is to get my post count up, that way I can be a regular and talk down to the newbies too.
Ummmmm thanks for that. I too, might take everything from my site one day and post it on the forums, just in a bid to get my word count up.......
'kinell!
Having missed all the excitment over the last few days, I return with a monstor post. This is largely my own work, but I have drawn upon many resources including the excellent http://www.searchenginewatch.com.
This is my first Masterclass, a thread where I write about something I think I know about, and where you guys get the chance to pick my brains if you like. This thread (and others to come if you want) serves 2 purposes to my mind, I share my knowledge with you, but you share yours with me too, that way, we all know a little bit more about the crazy web.
My first masterclass is on search engines, this is a report I first started writing about 3 years ago, but which has changed a few times since, and may be slightly out of date, but hey, you can't win 'em all.
Class begins.... this article is Copyright to me. It has been distributed to a few companies but contains nothing you cannot research for yourself on the Internet. If you blatently copy and paste it across the internet and fob it off as your own work you may find yourself in some hot water.
-----------------------------------------------------------
Copyright 2002 - All Rights Reserved
Contact [email protected] for permission to use this text, or portions of it anywhere else.
What is a Search Engine?
In the early days, when the web was just beginning, it was not hard to find a site that had what you wanted, as there were not many sites to look at in the first place. As the Internet (and therefore the web) grew, domain names were snapped up, and web sites became more and more commonplace. Now there was a plethora of information at the surfers' fingertips, the problem was not "is there a web site about widgets?" but was "where can I find a web site about widgets". Enter the Search Engines.
Essentially, a search engine is a database of web sites. Search engines allow users to interrogate the database for specific, or even non-specific information, the database should return a list of web sites based on the users' search criteria, the user can then select the web site they want and go merrily surfing off.
There are several different search engines, some of which are simply providing a different front-end (or interface) to a single back-end database. There are some search engines which do maintain a separate database of web sites but which may rely upon other services for data if their own is not sufficient to provide results to a search.
* Search engine - a big database which holds information on web sites, users can query the database for sites relevant to their needs. (http://www.northernlight.com is an example of a search engine)
* Search directory - like a search engine, the difference being that the directory is designed in such a way that users are supposed to "drill down" the directory structure (i.e., entertainment / tv / bbc1 / Monday / etc...) to reach the specific information that they want. (http://www.yahoo.com is the most famous search directory, but which is often mistaken for a search engine).
* Meta-search engine - type a query into a meta-search engine and it will send your request to multiple other search engines, the results of which will then be collated and displayed. A meta-search engine does not have its' own database, but relies upon others. (http://www.ask.co.uk for a good meta-search engine)
All of the above will be termed search engines from now on
Very few search engines exist to solely provide users with a search service. Many search engines exist to make money from banner adverts or other "services" or "buying opportunities" arranged around search results in such a way to distract the user from the results. It is a fair guess to assume that at least as much time money and effort is spent trying to distract users to banner ads, etc, as is spent searching out, collecting and categorizing web sites to put in the database. Google.com and dmoz.org are fairly unique search engines in that they exist simply to serve other search engines and do not have paid-for advertising.
How do search engines work?
As was mentioned before, there are many search engines out there and therefore many different methodologies are employed to:-
1. Find a site to register
2. Look at and "evaluate" the sites' worthiness to be included into the search engines' database.
3. Allow the user to search.
4. Display the site to the user after a search has pulled a specific web site from its' database.
1: Find a site to register
This first step is easy. Either a search engine uses a registration form for users to fill in and submit the site for evaluation, or the search engine employs "spiders". Spiders can trawl the web looking for sites independently of whether they are submitted for evaluation or not. Some search engines may combine both the form and spider technique, simply queuing up potential web sites that have been submitted via a form, until the spider gets around to looking at them ("spidering").
2: Evaluation of the web site
The second step is perhaps the most controversial. Once a site has been spidered, it must then be decided what category the site goes in (if this has not already been defined by a human submission form), to what extent Exactly how do search engines deem if a site is worthy or not to be indexed? There are many means at a search engines' disposal to decide upon this:-
* Meta-tags - there are 2 main types of meta-tags,
1. Keywords, in which you place the words that are most relevant to your site.
2. Description, which is usually a meaningful sentence about the web site. These are not visible to the average web surfer without looking at the source code to the web page.
* Document Title - every HTML document on your web site should have a title. This title is displayed in the blue bar across the top of a users' web browser.
* Body text - the visible text on the web page.
* Alt tags - ever held your mouse over a button or image for a few seconds, to see a little box pop-up with a description. Alt-tags are within the HTML source code of a web page but which define what the little pop-up boxes contain.
* Links to and from your site - spiders can follow the links that are defined within HTML to jump to other web pages and web sites altogether. Some spiders check to see if you link to sites already contained within their database, if you do, what category of site do you link to, what is the standing / popularity / quality of that site within the database. Similarly, a web site may be judged upon the web sites which link to it. If popular / high ranking sites link to a specific web site, and vice versa, the web site in question may be given a better position based upon this.
* Human evaluation - Sites like Yahoo! and http://dmoz.org/ rely upon humans (or "editors") to evaluate a site. If a submission is correctly filled out and the site does what it purports to do then (hopefully) the human editor will allow the site into the index.
* Domain name - what is your domain name? Does it match the theme of the site?
* Other points -
1. It has been suggested that some search engines actively use software to scan Internet newsgroups for relevant links to web sites
2. Once a spider (or editor) has visited a site, and the site matches the set criteria, this still does not mean that the site is placed in the search engine index immediately, it may be between 24 hours to six months before the site appears in the index at all, never mind near the "top" of the listing.
3. Some spiders revisit sites to ensure that they still exist, in doing so they may check the time at which the site was last updated. If the site has not been actively updated for an extended period of time it may be assumed that the data held on the site is less relevant or up-to-date than a freshly submitted site. The older site may be penalized and given a lower position in the index.
4. Many search engines like to publish that they have 20 million pages indexed and at the disposal of searchers, however quantity does not mean quality. How many times have you searched for "greenhouse" only to see "XXX rated" type websites appear in the results? (Yes, this has happened!)
5. Search engines may be using software to detect which web sites are being clicked onto the most from their listings. Those sites that are constantly avoided by users may be penalized.
6. The order in which keywords appear in meta-tags, title and body text has a bearing upon the weighting given to them, ie, words which appear first are deemed more important than words which appear after
7. The size of words in the body text has a bearing, large titles at the beginning of a document are given more precedence, and are sometimes used in the listings that search engines give.
With the exception of dmoz and yahoo!, it is estimated that differing search engines rely upon different combinations of the above to rank a web site. For instance, one search engine may simply take your meta-tags and title and from those decide upon where you should be placed, another search engine may place emphasis upon your meta-tags and body text, yet another may look at the amount of body text you have and the words contained within it, but will ignore the title to your document.
There is a phenomenal amount of calculations (or algorithms) which search engines can apply to web pages to work out if it gets into the index, and under what category. Educated guesses can be made as to which techniques a spider uses to index a web site, but unless search engines make public the exact way in which they decide how to rank a web site, it will be impossible to tell how it is done with a 100% degree of certainty.
In the early days, webmasters worked out that top rankings could be achieved by using "spamming" techniques such as using characters such as !, @, and others to fool the search engine into giving the site a higher alphabetical order, or hiding white coloured text on white background that contained multiple keywords to fool the search engine that enormous emphasis was being placed upon this subject. Whilst this has worked in the past, the top search engines have added anti-spamming techniques to their spiders, to the point where it is now not worth trying to spam a search engine as it will probably be detrimental to any listing, or even future attempt at a listing that a web site may have.
3: Allowing the user to search
Even if the first two stages are completed successfully, there is still the matter of web surfers being able to pinpoint a single web site amongst the many millions that can be held in the search engines index. Many users simply enter a few keywords, hit the search button and expect results, which is the simplest way of going about a search. This is akin to going into a library and saying to the librarian 'cooking' or 'gardening'. A more sophisticated searcher will start to use advanced Boolean logic functions available with some search engines that help to distill the database results even further. Instead of asking for "cooking", the user might ask for sites pertaining to "cooking" AND "Chinese", or they may use speech marks to denote that they want to view a site that contains information on "18th century Chinese cooking". Books have been written on the best ways of getting results from search engines, internet magazines are running tutorials on the how to use search engines effectively, so generally, public awareness about using advanced search functions is being raised.
4: Displaying Search Results
This final part of the effort is the part that should bear the fruits of fairly intensive labour. A common misconception about the results that search engines throw up is that there is a single list of web sites about a certain subject, and your specific site is either at the "top" or the "bottom". Search engines may display results in a manner of different ways:-
1. Alphabetically - A to Z, probably based on a sites' title or meta description.
2. Percentage based rating - sites that are deemed to closely match your search are given a higher percentage rating. Search engines like this are highly dynamic, applying algorithms to web pages in the database on-the-fly to come up with something (based upon meta-tags, title, body text etc) that it assumes is the best match to your search. Using this method, the same site may achieve wildly differing results for specific keywords. A site about called "cooking4u.com" may achieve really good results for "recipes" but not for the actual word "cooking". Both are still valid words used when searching for new dishes to eat. Again, the same site may achieve different results on a different search engines for the same keyword depending upon the internal calculations applied to the web site at the time of search.
3. Other internal search criteria - sites are not alphabetically listed, nor are they given a percentage mark, users just trust that the sites returned are relevant to the search conducted. It is likely that sites like these have their own version of the percentage based system as outlined above, but just do not show the final percentage that is given to the site.
The frustrating part of this is knowing that your web site has been spidered, and is contained within the index, but is simply not appearing for the keywords that you feel it should.
Best Practice
Given that we are aware of all the issues, how can we best promote a web site to the search engines? The first step is realizing that a "top" rating for specific keywords will not happen overnight, it will take a sustained assault upon the search engines to achieve anything like good results.
Upon first submission, sites may be indexed by search engines, but will almost certainly not perform too well except in a minority of cases. Perhaps after 6-12 months of constant re-submission and marketing the site will begin to achieve something like a high presence in the search engines. Think of it as market share. If you start a new company (or web site) up, you will have no market share, you must build market share up through hard work, and you cannot expect success to happen immediately upon start-up. So it is with web sites and search engines. It is also important to understand that achieving a good rating in search engines will not make or break a web site, search engines are but one piece of the web marketing jigsaw.
With this in mind, here are some search engine ratings tips:-
* Start thinking keywords at the design stage. Search engine marketing is not something to think about right at the end of the web site design and production cycle. Keywords should be defined and sprinkled about the web site pages during production.
* Only submit the site when it is fully operational, "under construction" signs are extremely undesirable.
* Is it possible to split the site into different sections, or sub-sites, (for instance, different products) and promote each section of the site individually to the search engines? Use different meta-tags and keywords for each unique section.
* Don't overdo the marketing slang. If you make caravans, use "caravans" as your keyword, not "holiday homes".
* Try not to get too technical - use words that the majority of people will know and use and understand.
* Get technical! Contrary to the above, if you are selling a highly specific product, to a specific group of people, ensure that you use the technical words that those people would use to find your product.
* Avoid ego trips - the Internet is famous for allowing local business reach a global audience, but the global audience probably does not know your business name. People tend too search using words like "playstation" or "gardening" or "caravan", they may not be looking for a specific company, but rather a specific product, so if you concentrate on pushing your corporate name onto the search engines, you are missing out on millions of people who do not know your company name, but who are looking to buy a product which you have.
* If possible, buy your company domain name as a .com and .co.uk domain. Many search engines give a higher priority to .com domain names (probably American bias, but also they expect .coms to be bigger and better then local regional domains.)
* If you market a range of products with trademark or individual names, try to buy the domain names of the products and point those domain names towards your main site (or sub-sites). This serves three distinct purposes.
1. Protect your trademark or product / intellectual copyright
2. The search engines will treat each new domain name as a separate website, so you stand a greater chance of appearing in the index through normal spidering activities. Your main domain name may not even fair as well as the other "satellite" domains for certain keywords, but does it matter as long as the site as a whole is doing well?
3. Many people randomly type in domain names into their browser, guessing that what they are looking for has a website under a specific domain. The chances of a site being found in this way are greatly increased if there are a number of relevant domain names pointing to it.
If you can purchase domain names that contain some of your keywords, this too can help to gain a better stronghold in the search engine indexes.
* Re-submit - about once a month make a habit of re-submitting the site to search engines. Contrary to what many search engines would like you to believe, there is a limit to how many web pages they can store in a database, eventually some web sites are removed from the index to make way for new listings. Constant re-submission means that every time the database is updated, you can be fairly sure that even if some of your pages are kicked out of the index, you have re-submitted the site so the spider should have allowed some more of your pages back into the index.
* Update the web site often - as mentioned before, spiders re-visit web sites to make sure they still exist, but also to ensure that updates are taking place. If a spider detects that a web site is getting dated it may decide to drop it from the index altogether. Sites should be updated regardless from a purely marketing point of view. Nobody wants to keep returning to a site with the same old information week after week. The truly great web sites are those which are never completed, but which are fluid and dynamic, and keep evolving week after week.
* Don't use frames, as spiders have difficulty with them. Those spiders that can cope with frames will probably index pages contained within the frames, but which will not make sense to a browser without the other frames around it to complete the picture.
* If the look and feel of the site is more important then anything else, use "bridge" pages instead of submitting your home page.
-------------------------------------
Copyright 2002 - All Rights Reserved
Contact [email protected] for permission to use this text, or portions of it anywhere else.
This is my first Masterclass, a thread where I write about something I think I know about, and where you guys get the chance to pick my brains if you like. This thread (and others to come if you want) serves 2 purposes to my mind, I share my knowledge with you, but you share yours with me too, that way, we all know a little bit more about the crazy web.
My first masterclass is on search engines, this is a report I first started writing about 3 years ago, but which has changed a few times since, and may be slightly out of date, but hey, you can't win 'em all.
Class begins.... this article is Copyright to me. It has been distributed to a few companies but contains nothing you cannot research for yourself on the Internet. If you blatently copy and paste it across the internet and fob it off as your own work you may find yourself in some hot water.
-----------------------------------------------------------
Copyright 2002 - All Rights Reserved
Contact [email protected] for permission to use this text, or portions of it anywhere else.
What is a Search Engine?
In the early days, when the web was just beginning, it was not hard to find a site that had what you wanted, as there were not many sites to look at in the first place. As the Internet (and therefore the web) grew, domain names were snapped up, and web sites became more and more commonplace. Now there was a plethora of information at the surfers' fingertips, the problem was not "is there a web site about widgets?" but was "where can I find a web site about widgets". Enter the Search Engines.
Essentially, a search engine is a database of web sites. Search engines allow users to interrogate the database for specific, or even non-specific information, the database should return a list of web sites based on the users' search criteria, the user can then select the web site they want and go merrily surfing off.
There are several different search engines, some of which are simply providing a different front-end (or interface) to a single back-end database. There are some search engines which do maintain a separate database of web sites but which may rely upon other services for data if their own is not sufficient to provide results to a search.
* Search engine - a big database which holds information on web sites, users can query the database for sites relevant to their needs. (http://www.northernlight.com is an example of a search engine)
* Search directory - like a search engine, the difference being that the directory is designed in such a way that users are supposed to "drill down" the directory structure (i.e., entertainment / tv / bbc1 / Monday / etc...) to reach the specific information that they want. (http://www.yahoo.com is the most famous search directory, but which is often mistaken for a search engine).
* Meta-search engine - type a query into a meta-search engine and it will send your request to multiple other search engines, the results of which will then be collated and displayed. A meta-search engine does not have its' own database, but relies upon others. (http://www.ask.co.uk for a good meta-search engine)
All of the above will be termed search engines from now on
Very few search engines exist to solely provide users with a search service. Many search engines exist to make money from banner adverts or other "services" or "buying opportunities" arranged around search results in such a way to distract the user from the results. It is a fair guess to assume that at least as much time money and effort is spent trying to distract users to banner ads, etc, as is spent searching out, collecting and categorizing web sites to put in the database. Google.com and dmoz.org are fairly unique search engines in that they exist simply to serve other search engines and do not have paid-for advertising.
How do search engines work?
As was mentioned before, there are many search engines out there and therefore many different methodologies are employed to:-
1. Find a site to register
2. Look at and "evaluate" the sites' worthiness to be included into the search engines' database.
3. Allow the user to search.
4. Display the site to the user after a search has pulled a specific web site from its' database.
1: Find a site to register
This first step is easy. Either a search engine uses a registration form for users to fill in and submit the site for evaluation, or the search engine employs "spiders". Spiders can trawl the web looking for sites independently of whether they are submitted for evaluation or not. Some search engines may combine both the form and spider technique, simply queuing up potential web sites that have been submitted via a form, until the spider gets around to looking at them ("spidering").
2: Evaluation of the web site
The second step is perhaps the most controversial. Once a site has been spidered, it must then be decided what category the site goes in (if this has not already been defined by a human submission form), to what extent Exactly how do search engines deem if a site is worthy or not to be indexed? There are many means at a search engines' disposal to decide upon this:-
* Meta-tags - there are 2 main types of meta-tags,
1. Keywords, in which you place the words that are most relevant to your site.
2. Description, which is usually a meaningful sentence about the web site. These are not visible to the average web surfer without looking at the source code to the web page.
* Document Title - every HTML document on your web site should have a title. This title is displayed in the blue bar across the top of a users' web browser.
* Body text - the visible text on the web page.
* Alt tags - ever held your mouse over a button or image for a few seconds, to see a little box pop-up with a description. Alt-tags are within the HTML source code of a web page but which define what the little pop-up boxes contain.
* Links to and from your site - spiders can follow the links that are defined within HTML to jump to other web pages and web sites altogether. Some spiders check to see if you link to sites already contained within their database, if you do, what category of site do you link to, what is the standing / popularity / quality of that site within the database. Similarly, a web site may be judged upon the web sites which link to it. If popular / high ranking sites link to a specific web site, and vice versa, the web site in question may be given a better position based upon this.
* Human evaluation - Sites like Yahoo! and http://dmoz.org/ rely upon humans (or "editors") to evaluate a site. If a submission is correctly filled out and the site does what it purports to do then (hopefully) the human editor will allow the site into the index.
* Domain name - what is your domain name? Does it match the theme of the site?
* Other points -
1. It has been suggested that some search engines actively use software to scan Internet newsgroups for relevant links to web sites
2. Once a spider (or editor) has visited a site, and the site matches the set criteria, this still does not mean that the site is placed in the search engine index immediately, it may be between 24 hours to six months before the site appears in the index at all, never mind near the "top" of the listing.
3. Some spiders revisit sites to ensure that they still exist, in doing so they may check the time at which the site was last updated. If the site has not been actively updated for an extended period of time it may be assumed that the data held on the site is less relevant or up-to-date than a freshly submitted site. The older site may be penalized and given a lower position in the index.
4. Many search engines like to publish that they have 20 million pages indexed and at the disposal of searchers, however quantity does not mean quality. How many times have you searched for "greenhouse" only to see "XXX rated" type websites appear in the results? (Yes, this has happened!)
5. Search engines may be using software to detect which web sites are being clicked onto the most from their listings. Those sites that are constantly avoided by users may be penalized.
6. The order in which keywords appear in meta-tags, title and body text has a bearing upon the weighting given to them, ie, words which appear first are deemed more important than words which appear after
7. The size of words in the body text has a bearing, large titles at the beginning of a document are given more precedence, and are sometimes used in the listings that search engines give.
With the exception of dmoz and yahoo!, it is estimated that differing search engines rely upon different combinations of the above to rank a web site. For instance, one search engine may simply take your meta-tags and title and from those decide upon where you should be placed, another search engine may place emphasis upon your meta-tags and body text, yet another may look at the amount of body text you have and the words contained within it, but will ignore the title to your document.
There is a phenomenal amount of calculations (or algorithms) which search engines can apply to web pages to work out if it gets into the index, and under what category. Educated guesses can be made as to which techniques a spider uses to index a web site, but unless search engines make public the exact way in which they decide how to rank a web site, it will be impossible to tell how it is done with a 100% degree of certainty.
In the early days, webmasters worked out that top rankings could be achieved by using "spamming" techniques such as using characters such as !, @, and others to fool the search engine into giving the site a higher alphabetical order, or hiding white coloured text on white background that contained multiple keywords to fool the search engine that enormous emphasis was being placed upon this subject. Whilst this has worked in the past, the top search engines have added anti-spamming techniques to their spiders, to the point where it is now not worth trying to spam a search engine as it will probably be detrimental to any listing, or even future attempt at a listing that a web site may have.
3: Allowing the user to search
Even if the first two stages are completed successfully, there is still the matter of web surfers being able to pinpoint a single web site amongst the many millions that can be held in the search engines index. Many users simply enter a few keywords, hit the search button and expect results, which is the simplest way of going about a search. This is akin to going into a library and saying to the librarian 'cooking' or 'gardening'. A more sophisticated searcher will start to use advanced Boolean logic functions available with some search engines that help to distill the database results even further. Instead of asking for "cooking", the user might ask for sites pertaining to "cooking" AND "Chinese", or they may use speech marks to denote that they want to view a site that contains information on "18th century Chinese cooking". Books have been written on the best ways of getting results from search engines, internet magazines are running tutorials on the how to use search engines effectively, so generally, public awareness about using advanced search functions is being raised.
4: Displaying Search Results
This final part of the effort is the part that should bear the fruits of fairly intensive labour. A common misconception about the results that search engines throw up is that there is a single list of web sites about a certain subject, and your specific site is either at the "top" or the "bottom". Search engines may display results in a manner of different ways:-
1. Alphabetically - A to Z, probably based on a sites' title or meta description.
2. Percentage based rating - sites that are deemed to closely match your search are given a higher percentage rating. Search engines like this are highly dynamic, applying algorithms to web pages in the database on-the-fly to come up with something (based upon meta-tags, title, body text etc) that it assumes is the best match to your search. Using this method, the same site may achieve wildly differing results for specific keywords. A site about called "cooking4u.com" may achieve really good results for "recipes" but not for the actual word "cooking". Both are still valid words used when searching for new dishes to eat. Again, the same site may achieve different results on a different search engines for the same keyword depending upon the internal calculations applied to the web site at the time of search.
3. Other internal search criteria - sites are not alphabetically listed, nor are they given a percentage mark, users just trust that the sites returned are relevant to the search conducted. It is likely that sites like these have their own version of the percentage based system as outlined above, but just do not show the final percentage that is given to the site.
The frustrating part of this is knowing that your web site has been spidered, and is contained within the index, but is simply not appearing for the keywords that you feel it should.
Best Practice
Given that we are aware of all the issues, how can we best promote a web site to the search engines? The first step is realizing that a "top" rating for specific keywords will not happen overnight, it will take a sustained assault upon the search engines to achieve anything like good results.
Upon first submission, sites may be indexed by search engines, but will almost certainly not perform too well except in a minority of cases. Perhaps after 6-12 months of constant re-submission and marketing the site will begin to achieve something like a high presence in the search engines. Think of it as market share. If you start a new company (or web site) up, you will have no market share, you must build market share up through hard work, and you cannot expect success to happen immediately upon start-up. So it is with web sites and search engines. It is also important to understand that achieving a good rating in search engines will not make or break a web site, search engines are but one piece of the web marketing jigsaw.
With this in mind, here are some search engine ratings tips:-
* Start thinking keywords at the design stage. Search engine marketing is not something to think about right at the end of the web site design and production cycle. Keywords should be defined and sprinkled about the web site pages during production.
* Only submit the site when it is fully operational, "under construction" signs are extremely undesirable.
* Is it possible to split the site into different sections, or sub-sites, (for instance, different products) and promote each section of the site individually to the search engines? Use different meta-tags and keywords for each unique section.
* Don't overdo the marketing slang. If you make caravans, use "caravans" as your keyword, not "holiday homes".
* Try not to get too technical - use words that the majority of people will know and use and understand.
* Get technical! Contrary to the above, if you are selling a highly specific product, to a specific group of people, ensure that you use the technical words that those people would use to find your product.
* Avoid ego trips - the Internet is famous for allowing local business reach a global audience, but the global audience probably does not know your business name. People tend too search using words like "playstation" or "gardening" or "caravan", they may not be looking for a specific company, but rather a specific product, so if you concentrate on pushing your corporate name onto the search engines, you are missing out on millions of people who do not know your company name, but who are looking to buy a product which you have.
* If possible, buy your company domain name as a .com and .co.uk domain. Many search engines give a higher priority to .com domain names (probably American bias, but also they expect .coms to be bigger and better then local regional domains.)
* If you market a range of products with trademark or individual names, try to buy the domain names of the products and point those domain names towards your main site (or sub-sites). This serves three distinct purposes.
1. Protect your trademark or product / intellectual copyright
2. The search engines will treat each new domain name as a separate website, so you stand a greater chance of appearing in the index through normal spidering activities. Your main domain name may not even fair as well as the other "satellite" domains for certain keywords, but does it matter as long as the site as a whole is doing well?
3. Many people randomly type in domain names into their browser, guessing that what they are looking for has a website under a specific domain. The chances of a site being found in this way are greatly increased if there are a number of relevant domain names pointing to it.
If you can purchase domain names that contain some of your keywords, this too can help to gain a better stronghold in the search engine indexes.
* Re-submit - about once a month make a habit of re-submitting the site to search engines. Contrary to what many search engines would like you to believe, there is a limit to how many web pages they can store in a database, eventually some web sites are removed from the index to make way for new listings. Constant re-submission means that every time the database is updated, you can be fairly sure that even if some of your pages are kicked out of the index, you have re-submitted the site so the spider should have allowed some more of your pages back into the index.
* Update the web site often - as mentioned before, spiders re-visit web sites to make sure they still exist, but also to ensure that updates are taking place. If a spider detects that a web site is getting dated it may decide to drop it from the index altogether. Sites should be updated regardless from a purely marketing point of view. Nobody wants to keep returning to a site with the same old information week after week. The truly great web sites are those which are never completed, but which are fluid and dynamic, and keep evolving week after week.
* Don't use frames, as spiders have difficulty with them. Those spiders that can cope with frames will probably index pages contained within the frames, but which will not make sense to a browser without the other frames around it to complete the picture.
* If the look and feel of the site is more important then anything else, use "bridge" pages instead of submitting your home page.
-------------------------------------
Copyright 2002 - All Rights Reserved
Contact [email protected] for permission to use this text, or portions of it anywhere else.