Controlling web traffic

The amount of traffic seen by a web site is a measure of its popularity. By analysing the statistics of visitors it is possible to see shortcomings of the site and look to improve those areas. It is also possible to increase (or, in some cases decrease) the popularity of a site and the number of people that visit it.

[]Limiting access

It is sometimes important to protect some parts of a site by password, allowing only authorized people to visit particular sections or pages.

Some site administrators have chosen to block their page to specific traffic, such as by geographic location. The re-election campaign site forU.S. President George W. Bush (GeorgeWBush.com) was blocked to all internet users outside of the U.S. on 25 October 2004 after a reported attack on the site.^[1]

It is also possible to limit access to a web server both based on the number of connections and by the bandwidth expended by each connection. On Apache HTTP servers, this is accomplished by the limitipconn module and others.

[]Increase web site traffic

Web traffic can be increased by placement of a site in search engines and purchase of advertising, including bulk e-mail, pop-up ads, and in-page advertisements. Web traffic can also be increased by purchasing non-internet based advertising.

If a web page is not listed in the first pages of any search, the odds of someone finding it diminishes greatly (especially if there is other competition on the first page). Very few people go past the first page, and the percentage that go to subsequent pages is substantially lower. Consequently, getting proper placement on search engines is as important as the web site itself.

[]Organic traffic

Web traffic which comes from unpaid listing at search engines or directories is commonly known as "organic" traffic. Organic traffic can be generated or increased by including the web site in directories, search engines, guides (such as yellow pages and restaurant guides) and award sites.

In most cases the best way to increase web traffic is to register it with the major search engines. Just registering does not guarantee traffic, as search engines work by "crawling" registered web sites. These crawling programs (crawlers) are also known as "spiders" or "robots". Crawlers start at the registered home page, and usually follow the hyperlinks it finds, to get to pages inside the web site (internal links). Crawlers start gathering information about those pages and storing it and indexing it in the search engine database. In every case, they index the page URLand the page title. In most cases they also index the web page header (meta tag) and a certain amount of the text of the page. Then, when a search engine user looks for a particular word or phrase, the search engine looks into the database and produces the results, usually sorted by relevance according to the search engine algorithms.

Usually, the top organic result gets most of the clicks from web users. According to some studies^{[citation needed]}, the top result gets between 5% and 10% of the clicks. Each subsequent result gets between 30% and 60% of the clicks of the previous one. This indicates that it is important to appear in the top results. There are some companies which specialize in search engine marketing. However, it is becoming common for webmasters to get approached by "boiler-room" companies with no real knowledge of how to get results. As opposed to pay-per-click, search engine marketing is usually paid monthly or annually, and most search engine companies cannot promise specific results for what is paid to them.

Because of the huge amount of information available on the web, crawlers might take days, weeks or months to complete review and index all the pages they find. Google, for example, as of the end of 2004 had indexed over eight billion pages. Even having hundreds or thousands of servers working on the spidering of pages, a complete reindexing takes its time. That is why some pages recently updated in certain web sites are not immediately found when doing searches on search engines.

[]Traffic overload

Too much web traffic can dramatically slow down or even prevent all access to a web site. This is caused by more file requests going to the server than it can handle and may be an intentional attack on the site or simply caused by over-popularity. Large scale web sites with numerous servers can often cope with the traffic required and it is more likely that smaller services are affected by traffic overload.

[]Denial of service attacks

Denial-of-service attacks (DoS attacks) have forced web sites to close after a malicious attack, flooding the site with more requests than it could cope with. Viruses have also been used to co-ordinate large scale distributed denial-of-service attacks.

[]Sudden popularity

A sudden burst of publicity may accidentally cause a web traffic overload. A news item in the media, a quickly propagating email, or a link from a popular site may cause such a boost in visitors (sometimes called Slashdot effect or Digg or Reddit Effect) that overwhelms the site.

Web sites have been forced to close after an unexpected mass increase of traffic, particularly those run by an individual leasing the bandwidth from an ISP or hosting site. Some sites backed by large companies running their own servers have also been caught out by the problems of overpopularity. When first announced, the Vision of Britain Through Time site, containing information taken from the 1901 UK census, was advertised on numerous television programmes and causing such interest that the site had to be taken offline until different arrangements were made to cope with the traffic. The site was hosted by a project at the University of Edinburgh and they had not foreseen the amount of bandwidth and the server load that would be required. Ironically, by the time the site was able to cope with the traffic both the interest and the free advertisements of the site had greatly slowed, giving them excess capacity.