Please note: Every link on this page opens in a new window. If your "Pop-up killer" is too efficient it can also stop new windows. When this happens, please press "Ctrl" and click on the link you want.

Definitions

Paradigm

A paradigm is a mathematical concept, a set of rules telling you how to solve a certain kind of problem. In connection with search engines it is a program consisting of a set of rules on:
- how to search,
- how to analyse pages found,
- how and what to classify, and
- how to index internet pages found, analysed, and classified.
Every search engine uses a different search paradigm.

Algorithm

An algorithm is also a mathematical concept, giving a set of rules that must be followed when solving a certain problem. For instance how to calculate the ranking of a certain internet page in the search result presented.

What has This to do with Marketing?

Search engine marketing is about influencing (without cheating) the search engines' automated programs, i.e. the search paradigms and the ranking algorithms (see definitions above).

To successfully use search engines as media you need something from them:

First: You want your webpage to be chosen over others so it's placed as close to the number one slot as possible.

Second: You want the search engines to describe your webpage as invitingly as possible.

Third: You want the description on the search results page to be not only inviting but also correct and relevant to the search query.

To achieve this you try to exercise as much control as possible. How?

The search engine's business is to present relevant information to the person searching for something. You must in your turn present this to the search engines in such a way they recognize it as relevant for a certain search query. This process is sometimes called "search engine marketing" and sometimes "search engine optimization".

Once you have achieved an acceptable visibility and presence on the search engine's results pages ( SERP ) you give the search engines the words and expressions that will convince the searcher to visit your webpage.

That's what this webpage - including the "related pages" - is about.

How not to Get Banned

Get Google's advice and instructions on search engine optimization and webmaster guidelines. If you get banned there's no excuse because guidelines are publicly available.

One sure way to get banned, is to use so called "black-hat" techniques. These are tricks that aim to cheat Google and other search engines to give a website better ranking, without the webmaster or designer doing any much work for it. If you suspect you are unwittingly in the risk zone, check these points:

• Are there duplicate pages. If you have "printer-friendly" versions, use robots.txt file.
  
• Do you use re-directs? These can be used for cloaking.
  
• Do you employ keyword stuffing? Keep your text natural sounding, if it sounds unnatural you may be have too many keywords (max 1-2 %) or too many repeated keyword phrases - each one more than 4 times per webpage ?

End of 2006 Live Search started penalizing websites for exchanging unrelated reciprocal links, for instance via link farms or links exchanges.

If you have a suspicion your website has by some reason been banned, you can check if a website is banned or not. It doesn't work with a very new website.

If your site's banned, then after you have corrected the bad mistakes you can ask for re-inclusion in Google by using the form (Google Webmaster Tools). In MSN Live you can also ask for re-inclusion.

What is a search engine ?

Search engines are the medium that carries your message to the persons you are targeting. Therefore you need to understand how the search engines work before you can make a webpage that is attractive to them. The content (text and graphics) you make for the visitors but you have to get the search engines to choose to present the webpage to your customers.

There are two kinds of search engines:
- Regular ones, crawling the net, registering pages, and/or admitting submissions from the public.
- "Meta" search engines, collecting their search results from the "regular" ones.

If the pages are not classified and indexed nobody would be able to find what they are looking for. In most cases the classification involves text analysis. Your page content will in itself decide how it will be classified. As the "robot" picks out "keywords" from your text, the amount and variation of these will also influence the ranking of your page on the results pages. Will it be on the 1^st or on the 15^th results page?

The search engine "spiders" ( programs ) are reading your code as written. Errors in your code ( especially your links ) as well as spelling errors are negative factors, which will reduce your site's ranking in the search results.  Only a few surfers check the results for more than 2-3 pages.  So if your site is on the 15^th results page, you end up with 0 = zero!

Focus your text on the visitor, not on the search engine. The search engines are the medium carrying your message to your target group. The code you write for them especially.

Meta tags or no tags ?

On different discussion forums you can still find an on and off "quarrel" between "experts" whether one should use meta tags or not. Especially the tags "Keyword" and "Description". The usual reason given for leaving these out, is that the search engines do not use them anymore.

It's true many search engines don't read meta tags, because of earlier misuse by designers. Instead these search engines pick the keywords from the text and the description part from one or two sentences in the beginning of the text. However, I know for a fact, that at least some search engines do use the "Description" tag as it's written. This I know, because I have seen the description of some of my pages in the search result, word by word exactly as I have written the description tag on those pages.

More on meta tags in Don Pedro's Meta Tags and Search Engines.

Search Engine Usage

Below is given statistics for search engine usage and traffic sources from 2007 to 2009 for the Global Maritime Sector.

The regional versions of each search engine has been lumped together respectively.

The percentage of traffic from incoming links went down in 2007 because I moved the website and changed domain name. Several webmasters never bothered to change the links to my website so I lost a few links and some traffic from those. Still more than two years after moving the domain the old domain is still sending a steady 50-60 % of the visitors forward to the new location. There is therefore good reason to leave the old location running on a free server, just linking to the new location.

On the other hand, traffic from search engines multiplied, while traffic from incoming links developed more slowly. This continued in 2008 - 2009 and incoming links couldn't keep up with the increase of search engine traffic. Starting about February 2009 the increasing unemployment and especially the continuous news in news papers and TV about unemployment in North America and Europe did have an dampening effect and traffic slowed down. This development continued until the end of 2009 to change strongly to an upward trend in Nov. / Dec. 2009.

Google got a high percentage in 2007 because I moved the website from US to UK. Google's market share is considerably higher in Europe than in US, while Yahoo and MSN + Live Search have a higher market share in US. In June 2009 MSN scrapped Live Search and introduced a new search engine: Bing. It didn't have any much effect in the short time in June, Bing showed up at 0.2 % only. At the end of the year Bing had picked up the traffic from old Live Search and was unchanged. The new deal between Yahoo and Microsoft ( "Bingahoo" ) took first a market share of 5.6 % when people tried the new comer. This dropped at the end of the year and both Yahoo and Bing suffered.

Google's age distribution for their users is very close to normal distribution ( i.e. "bell shaped" ), while Yahoo's audience is skewed towards younger people ( 18-24 years old ). Then again, the target for that maritime employment portal is the mature professionals, that's is roughly from 25 years old and up.

The percentage of returning visitors is a calculated value of visitors who have either bookmarked or typed the URL directly. "Navigational queries" are included in "returning visitors". I don't like to overstate, so the above values are conservative.

Always must remember website statistics are website specific. This is clear when comparing the search engine market shares for 2008 in above table ( global audience ) with search engine market shares in USA in December 2008:
- Google 63.5 %,
- Yahoo 20.5 %,
- MSN / Live 8.3%,
- Ask 3.9 %, and
- AOL 3.8 %. ( powered by Google )

There are, three different kinds of statistics. What the search engines publish is naturally from their viewpoint. They need to know how the total number of "searchers" is divided between the search engines. The figures for the Maritime sector above are from the site owner's view = "How is the traffic from search engines divided between each of them?". Something completely different!

The third kind of statistics is from the "searcher's" viewpoint. "What search engines are others using?"  You can get an approximation of this at International Search Engine Statistics (English, Español).

MSN continues to have problems with the relevance of the search results they give. That's the reason why Google is so high with MSN very low - when looking at averages for the whole year. End of 2007 it seemed Live search started getting their act somewhat together. But this soon dropped off.

How do People Search ?

Americans and Europeans search in different ways. American men are mostly searching for specific details; like hobbies, less than one third searching for job and career info. American women did more search about health and medical problems. Europeans did not search so much for hobbies and Google's share of searches was smaller than in USA (in %).

In 2007 Google's market share in Europe grew stronger than in US. Still local European search engines were generally number two. Possibly part of Google's success in Europe ( report from Aug. 17, 2007 ) is because Google has adapted their local versions to local languages.

During weekdays (2006) about 70% of Scandinavian (Denmark, Finland, Norway, and Sweden) users got online at least once, while just 55% of the British did so.

There is an American report on search engine users' behaviour when searching (valid for US online population). About 62% look only at the first page and 90% stop at the third results page. The trend to move after 1^st and 3^rd page is getting stronger, although 41% add one or several keywords to the search term before moving. In other words, you should use long key phrases (4-5 words) to get targeted searchers to your website.

A Dutch eye tracking report     exploring differences in eye movement patterns between people searching for information and those searching for products states keywords which refer to a brand are very important to induce people to click on a certain search result. This would indicate there's strong reason to include Company or Brand name in the title tag.

Further it was found consumers searching the Internet in order to make a transaction view a larger number of search results and are more influenced by branding than consumers searching for information.

The average viewing time per search result was only one second. In this time the searcher decides whether to click on your link or some other one of the search results on the search results page he/she is viewing.

Generally people do know very well what they want to find but they can have problems expressing this for the search engines. Especially it's so for non-native English speakers searching in English. The result is sometimes you get in your traffic data keywords that for you looks like "nonsense". This is so because you have no idea what was in the searchers mind when he/she typed in the search query.

An other likely reason could be, the searchers never search for keywords, he/she searches for information. Very often the reason for the search is the person is unfamiliar with the subject and therefore needs to know something. Anyone who is unfamiliar with something will easily ask for something that sounds silly to the professional, isn't that so?

What is a Good Search Engine?

A good search engine is the one where you find your information fast and easily.  What is good for me, that is may be only third rate for you.

Because the search paradigm for every search engine is different the search results are also different. The paradigms themselves are an "industry secret" because of possible misuse. The result is that some are good in one area and others are good in other areas.

Google seems generally to be stronger with technically oriented subjects and products, while Yahoo still is the general department store ("jack-of-all-trades"). Google's share of business and professional search is generally greater than Yahoo's, except in Europe the difference between Google's and Yahoo's shares is smaller than in USA.

If you try a certain string of search words (2 or 3 words) in several search engines you will notice big differences in the search results pages. Usually no more than two web pages will appear on two or more search engines, the rest can be wildly different. Comparing results from 2005 and 2007 it's clear this difference has increased. See Don Pedro's List of Search Engines.

What is Google's Caffeine Update ?

It's not an algorithm update. It is a differently organized index of the documents Google has found and stored. The basic idea seems to be to give more weight to fresh content - social media ?. You can test the US based beta index at:
http://www2.sandbox.google.com/
End of July 2009 Google opened a preview of results from their "Caffeine" index update. It's said that some results will be different but I didn't notice any big differences for my site. On the other hand, I'm up-dating at least some pages every week so the material is pretty fresh.

Because the "Caffeine index" is US based ( American English ), it's somewhat difficult to get any comparisons for other countries / languages. Google Webmaster Central Blog gives some advice on how to manually change the search parameters to give results for other countries / languages. When / if you try those, keep in mind the index is still US based. Webmasters and developers are invited to give feedback / comments on results found.

When first searching on regular google.com and then on "caffeine" it can be difficult to notice differences. You can, however, use "facesaerch's" caffeine setup for such a comparison [ Note the spelling: saerch !]. This gives you regular google results and caffeine results side by side on the same webpage.

Middle of November 2009 Caffeine started to roll, data center by data center and the "sandbox" was retired ( closed down ). In connection with this Google has started to incorporate download speed / time in the ranking algorithm.

What's new in Bing ?

First of all Bing started from the very beginning to target "travel and shopping". These are of course the very heavy part of total number of searches - in addition weight has been added also to "finance" related queries.

Like Google Bing looks at backlinks ( incoming links ) quite a lot. But Bing seems to take back links as more important than page content. May be Bing hasn't got the content analysis right yet ?

As in most search engines domain age is important as well as original content on each page. Page titles are very important also in Bing. Of course, every page must have a different <title></title> tag - every page is different, isn't it ?.

Starting middle of November 2009, outside of USA and UK only a beta version is available. This one has none of the additional features the American and the UK versions has. To get the American ( and UK ) version go to Search Engine List. At the bottom of the page there's advice how to do that.

Links and more links - Bing is said to love outgoing links. I have always on many of my pages said: "be generous with your outgoing links". The Internet is all about sharing information, isn't it ? That seems to work well in Bing. See Pandia for detailed review.

Java- and other Scripts

If you already are using, or if you are planning to use, some (one or more) JavaScript(s) or some other script(s) on your page(s), be careful. Don't use too many! In Dec. 2006 about 6% of visitors to this website had JavaScript switched off.

The search engines read HTML only, and they read it exactly as the code is written. These are computer programs not humans, so they have problems "understanding" lay-out tables (especially "nested tables" with some content) as well as Java and other scripts.

Especially you should avoid writing your Menu in JavaScript, because if you do, the search engines will have problems finding your pages. And if they don't find them, the pages won't be registered either.

With style sheets (CSS), you have to put your style guidelines into a separate file ".css" on your server. Then you call up this file with a call-up code in the <HEAD> section of your code. If you are using a free or a low-cost server, then it's possible it doesn't work. So may be better avoid to use CSS.  See Don Pedro's Cascading Style Sheets - Pro and Con.

If you put your "styles" in the <HEAD> section, the banner on a free server will cut the style descriptions into half - same problem as above.  You can, of course, put CSS as separate tags for every paragraph, but it makes your code longer and more complicated.  Simple HTML is easier to write and faster to download.

For JavaScripts you can do what is said about CSS above. Only difference is you have then one or several ".js" files on your server. These also require a call-up code in the <HEAD> section.  External scripts increase the down load time for your pages.

The search engines read HTML only.  If some other code isn't absolutely necessary - don't use it.

Dynamically Generated Sites

This is about DHTML. The bigger search engines' spiders are indexing some dynamic pages - i.e. most but not all of them. If you use "Session IDs" in your URLs, then those will still (2007) be avoided. In Oct. 2006 Google changed their guidelines for webmasters, still you are recommended to avoid URLs with "&id=".

You can modify your dynamic URLs so they look like static ones.

A dynamic page is something that's "made up on the fly" - drawn up by the browser from bits and pieces in a data bank. For instance because of a search query or when a visitor fills in a form with: model, colour, size etc. That gives an unique page that exists only as long as that visitor views it. When the window is closed that page doesn't exist anymore. The search engines cannot fill in forms so they cannot create dynamic pages and of course they cannot register non-existing pages either.

A static page is stored on the server in the same way it's presented both to spiders and visitors, which is what the search engines want.

It's a double faced problem. The first one is the URL - every detail delivered from the server has its own URL.  So a site with 100 pages only, can easily produce 10.000 or more URL's - with the single details changing, may be several times per hour.  The other problem is the form of the URL.  It includes a string of symbols not included in standard HTML, like: "?, &, %, +".

If you still want Dynamic pages, or may be the content details on your pages are changing daily, there is only one fully reliable way. You make "copies" of your DHTML pages in simple HTML and submit these to the search engines. Of course, your DHTML pages would then receive a "no index" tag. That is so you are not accused of using "double" pages and that way trying to cheat the search engines to give your website or webpage a better ranking.

Then you link from your HTML page so your visitor can get directly to the continuously up-dated DHTML page. If you make that link in an encrypted form, then you don't confuse the search engines. Arrange so that your incoming links point to your HTML pages only.
More info on dynamic pages.

Flash

The major search engines read and register Flash pages with the help of a "Text extraction" program developed by Macromedia.

This doesn't mean it's OK to start using Flash everywhere. If you are selling Flash, then, of course, you have to use it - it's your product! I can also think of some photo studios and art galleries who could benefit from using Flash (sparingly).

I happened to come to one site, which boldly stated "Site requirements: Flash and 1024 x 768 resolution"!! Isn't that like asking for an entrance fee? I wonder how many people never got any further than the first page.

If you, for instance, would be selling a new tool that is used in a completely new way - then may be you could use Flash to show how to use the tool. But even in this case, animated GIF pictures would probably download faster and be as efficient. See Don Pedro's How to Optimize Flash Pages for suggestion how to add optimized text to your Flash movie.

Before using Flash be absolutely sure the Flash will really benefit your visitors. Flash can be an effective tool when used intelligently and sparingly.

When you choose between Flash and text; main concern is to keep it simple. Text gives much faster download and display. Do your visitors want Flash? Put one counter on your Flash "entrance" page and another on "skip to" page; the difference minus those visitors who have entered the "skip to" page directly is the number of visitors lost.

The biggest problems with using Flash comes from the fact that there's new versions of the "Flash player" about every 18 months. Very few Internet users upgrade their Flash player all the time. In that case may be a visitor still cannot see anything of your beautiful Flash page, because the visitor's version is older than what's been used to create the webpage. Again more visitors lost.

If you are sure it's absolutely necessary to use Flash pages, don't try to optimize the text on those pages too much. To achieve a decent ranking in the search engines' results pages also for these pages concentrate on optimizing the anchor text together with the relevance of your incoming links.

Ask other webmasters not to point the incoming links to your Flash entry page. Get the links pointed to a regular static text page instead. Use a robots.txt file to exclude ("Disallow") your Flash entry page from the search engines.

What is a "Silo" ?

"Silo" is a new term some search engine marketers ( SEM ) have started applying for a slightly different kind of search engine optimization (SEO). Basically it means that within your website you build one or several separate networks of webpages - each network optimized for one or a few keywords or key phrases.

Why is this then called a silo ? The network is built up with internal links in such a way that each "silo" consists of a specific theme or sub-topic. The search engines require at least 5-6 pages - each 1-2 print-out pages text - to be able to identify the network as such and also to be able to classify the network (silo) as covering a specific subject (theme).

The network can be placed in it's own directory (folder) or the pages can be spread over different folders depending on your website technical information structure. When the pages are spread over several directories the network is sometimes called a "virtual silo". When the pages are placed in the same directory there's really nothing new about the idea - just a new name only.

Whether the silo is a "virtual" one or not, how you build the network links depends on what kind of menu you are using. Remember to keep the menu as simple as possible, you can easily loose what you gain by "siloing" if your menu is too complicated for the search engines to read and follow.
Compare for instance "Related Pages" at the bottom of this page.

Website Popularity

Teoma, Direct Hit, and Alexa have long been employing traffic as a factor when calculating page ranking in the search result. I have a feeling this is getting important also with the major search engines.

Click popularity registers how many people click at each site in the search result. There can also be a "stickiness" measurement, which registers how long a visitor stays on a site. This can be measured because when you return to the search results page it will be registered as "reload" by the search engine's server.

The idea behind it is that when a visitor goes to one of the first sites, if it's no good he/she stays only a short time and then goes on to the next one, etc. When he finally finds what he's been looking for he will probably stay longer. That means the content on the last site must be what he was looking for. Accordingly that page will get "plus points" and a higher ranking, while the earlier pages that were not so good will get "minus points" and a lesser ranking. There is furthermore a tendency among search engines to include accessibility in their ranking algorithm together with popularity.