Analysts have been trying to find answers to this one question ever since 1990s - what is the total number of pages internet hosting right now? The answer is a strict "INFINITE". Fact is, internet or internet hosting is not owned by a single person or organization, so we would never know the actual strength. Second, many web pages or web sites have auto generated URLs or content so if you include them, the number of web pages will run into "Infinite".

Still, we can apply assumptions, do some testing and estimate a plausible enough number of world wide web. From estimates of 220 million in 1990s, we have arrived at the latest number of 1 Trillion - not bad for a period of 10 years. Her, I am providing three estimates of the number of total web pages on internet. All three have their own flaws, but these are the only estimates you would get anywhere, so be content with them.

By the way, there is another point to this whole question - invisible internet which is not indexed or not indexable by search engines is at least 100 times larger than the visible internet which gets indexed in the search engine. Here we are talking only about the visible internet.

The first estimate

In a recent blog, a Google engineer announced that Google had seen a new landmark of 1 Trillion unique URLs. This number stands after eliminating duplicate entries. Although, some of these are auto generated content (e.g. a web calendar could have infinite links for each new day. If you start counting these, the number would turn into infinite), this is the best available number from as they say "the horse's mouth". Here is the link to his blog.

The second estimate

A private researcher has also been tracking the number of web pages on the basis of his own methodology for last two years. According to him, internet has at least 27 billion web pages and could have as many as 58 billion web pages. His methodology based on yahoo and google search results has its own flaws but these are the best organized estimates available in want of any other credible research. Problem is, Google based number (27 billion) sounds too small ( I will tell you why) and yahoo based numbers (58 billion) can not be taken seriously because yahoo search results do not omit duplicate entries. You can have a look at his research findings on this link.

The third estimate

I tried corraborating these two numbers and came up with a number of 48.27 billion. However, this is just a back of envelop calculation. Here is how I did it.

I searched a number of most common words and characters on Google (e.g. is, the, are, 0,1,2 etc) and found that Google search returned over 25.3 billion web pages for the search key word '1' - the highest among all the key words I used - probably because it is also used in many non -english web sites in the same form.

Of course not all the web pages have '1' and unfortunatly there is no way of knowing the percentage of such web pages. Very liberally, I assumed that one third or 33.3% of documents/web pages would not have '1' anywhere in them because of a different language or because they simply did not need to use '1'. Adding these back, we have 33.8 billion web pages indexed by Google alone. In 2005, Google claimed to have indexed 70% of visible internet. Assuming Google has not been able to increase this number, I have the total number of web pages at web pages indexed by Google divided by 0.7 i.e. 48.27 billion pages.

I would like to think that the 1 Trillion number would have many auto-generated content links and therefore I would be tempted not to trust the number entirely. Still, it does give an indication that the size of visible internet is bound to be quite big and definitely around 50 billion if not more than this. The other methodology explained is flawed as well. Google search returned 25 billion plus web pages for a single keyword ('1') whereas that guy estimates the size of internet at 27 billion. Not possible is an understatement here.

So, here you are - with three numbers. Take your pick.