![[OlympusNet Logo]](http://www.olympus.net/ONet_images/ONetLogo_sm.jpg)
|
COLUMN ON PUBLICITY FROM THE FINANCIAL TIMES |
|
|
COLUMN ON PUBLICITY FROM THE FINANCIAL TIMES |
|
ABOUT THIS COLUMN This is a copy of the a column that appears every Monday in the Financial Times newspaper. TIM JACKSON'S WEEKLY COLUMN So you've got the company's web site up and running. Now you face the difficult question of how to get people to come to it. Most web site owners simply assume that their domain name, such as www.sun.com, will bring in existing customers who are looking for it. But as more names are registered, each web site has to compete with more confusingly similar competitors. Advertising, often seen as a panacea, can be an expensive mistake. A recent column in this slot looked at the pitfalls of indiscriminate banner advertising on the Internet. In fact, the most important way of bringing people to a web site is also usually the most overlooked: the search engines, including Lycos, Altavista, WebCrawler, Excite, and HotBot. Many users start their research by typing in a string of words to one of these engines, and following the list of most relevant links that come back. Yahoo!, the most popular, delivered more than one billion page views in April 1997, and serves roughly 10% of the entire world's internet population. With listings far larger than the biggest library catalogues, the search engines are wonders of the modern world. They send out software 'spiders' and 'robots' to roam the web, returning with information on sites and what they contain. But these spiders and robots aren't human. They follow rigid rules in deciding what to index and how to rank it, and these rules have a dramatic effect on how each web site is indexed. If a web site is designed without giving any thought to how search engines work, a web user who types in a word or phrase that captures the site's subject perfectly is unlikely to see it. Among the thousands or hundreds of thousands of pages returned by the search, it will be buried too far down the list for any but the most persistent customer to see. Generally, what matters is to become one of the top ten 'hits' returned by the search. In recent weeks, I've received half a dozen pieces of junk mail offering to explain, for sums ranging from $20 to $100, the secrets of how to get your site into the Top Ten. So as a public service, I've spoken to some technical people at the search services and come up with some free advice on the matter. Step one, say the search people, is to keep your web pages pithy and relevant, so as to avoid confusing the robots with material that customers are unlikely to be searching for. The same goes for the title of each web page - the phrase that appears at the top of the browser when it arrives on the client's screen. What matters is not only the total number of mentions of a given word, but also how frequently the word appears as a percentage of the total document length. Step two is to use meta-tags. These are nuggets of hypertext markup language (HTML), designed specifically for search engines, which give a description of the page or a list of keywords under which it should be indexed. Forgetting to include these tags is probably the most common webmaster's mistake. Another mistake is to forget that robots and spiders are blind: they can't read graphics files, so they cannot decipher words that appear in fancy graphical fonts. In choosing meta-tags, it's helpful to be comprehensive. An agency that rents houses might include these keywords: realtor, house, flat, rental, new home, apartment, estate agent, moving, huose. Note the mis-spelling of the last word; this is deliberate, to try to attract the attention of Internet users who make a typing mistake during their search. Some webmasters, known in the trade as 'spandexers' because they are trying to spam the indexes, also use black arts. They repeat the same keyword repeatedly, or they include material at the bottom of the page in tiny white letters on a white background. After achieving a high ranking, some also design multiple identical pages to squeeze out the competition from the top ten. For instance, I found one seller of microprocessors whose many pages all contained the word 'MMX' 174 times and the word 'computer' 50 times. The search engines are getting wiser. Like taxmen on the trail of accountants, they keep tabs on what the spandexers are up to, and change their relevance ranking programs to punish the miscreants. But there are still techniques open to the honest webmaster. One is to try the kind of search that your own potential customers might carry out, and check the source code of the sites that come up tops. Another is to change the titles of pages and continually resubmit them. A third is to use the spandexers' own weapons against them. Since the search engines only started penalizing manipulation in late 1996, you can 'resubmit' someone else's web site to the search engine. This will prompt the spiders and robots to take a second look - and to delist a site if they find evidence of wrongdoing. I found three sites particularly useful on this topic. One is deadlock.com/promote, which contains good advice from Jim Rhodes, who explains how he got his London hotel to the top of the rankings, and in the process brought in dozens of new email bookings each day. Another is searchenginewatch.com, which contains detailed technical background and suggestions. A third is rankthis.com - a free service which allows you to type in a search term and the address of your own web site, and find out how highly you ranked. Unfortunately, the game is zero-sum, since not everyone can be in the top ten. So all advice, no matter how good, suffers from the same problem: once everyone else follows it, you're back to square one. ...................................................... ABOUT THE FINANCIAL TIMES The FT is an international business newspaper, edited in London and printed at numerous sites all over the world. It is available by hand delivery and subscription in many countries of the world. See http://www.ft.com for details. A selection of articles from the paper can be found on the web site, but not an archive of back issues of this column. ABOUT TIM JACKSON Tim Jackson, 32, writes a weekly column for the FT on the computer industry, the Internet and telecommunications. He was formerly on the editorial staff of The Economist and of The Independent newspaper, a national British daily paper, serving at various times as business writer, feature writer, leader writer, and foreign correspondent in Tokyo and Brussels. He is the author of three books:
Tim Jackson |
Updated 4/6/99