16% of searches every day - new
In one of his presentations, Google has published an interesting figure . It turns out that every day a search site processes 16% of new search queries that have never been seen before.
It would seem, where they come from, if all the words from the dictionary with all possible typos and in all combinations are already exhausted? But if you think about it, you can easily understand where they come from. First, a couple of hundred languages ​​other than English are contributing. Secondly, users very often search for specific information in long unique search queries (addresses of specific institutions, names and biographies, quotes from songs, literary works and code snippets, and much more). In the end, these are meaningless keyboard shortcuts that accidentally hit the search engine, as well as incomplete queries from a “dynamic” search that updates the results as the user types.
Perhaps, in statistics not the aggregate volume of requests is taken into account, but individual requests. For example, if a million people were searching for [facebook], and another person was searching for [dfg8734kjv3d], then in such statistics 50% of requests will be “completely new”. That is, from the average daily volume of 3 billion search queries processed by Google, not at all 480 billion are completely new, but an order of magnitude less.
Anyway, it turns out that for a certain period of time (several months?) The proportion of unique search queries that met only once may already exceed the number of “popular” search queries that have been met two or more times.
')
If we assume that the share of unique requests in the total volume is only 1%, while 90% of them are never repeated in the future, then it turns out that unique “one-time” requests reach 50% in the list of requests in 55 days and reach the level of 99, (9)% in the list of requests for 110 days of the search engine.
Even if we reduce the share of new requests to 0.1%, it’s still the milestone of 99, (9)% comes amazingly fast.
It would seem, where they come from, if all the words from the dictionary with all possible typos and in all combinations are already exhausted? But if you think about it, you can easily understand where they come from. First, a couple of hundred languages ​​other than English are contributing. Secondly, users very often search for specific information in long unique search queries (addresses of specific institutions, names and biographies, quotes from songs, literary works and code snippets, and much more). In the end, these are meaningless keyboard shortcuts that accidentally hit the search engine, as well as incomplete queries from a “dynamic” search that updates the results as the user types.
Perhaps, in statistics not the aggregate volume of requests is taken into account, but individual requests. For example, if a million people were searching for [facebook], and another person was searching for [dfg8734kjv3d], then in such statistics 50% of requests will be “completely new”. That is, from the average daily volume of 3 billion search queries processed by Google, not at all 480 billion are completely new, but an order of magnitude less.
Anyway, it turns out that for a certain period of time (several months?) The proportion of unique search queries that met only once may already exceed the number of “popular” search queries that have been met two or more times.
')
If we assume that the share of unique requests in the total volume is only 1%, while 90% of them are never repeated in the future, then it turns out that unique “one-time” requests reach 50% in the list of requests in 55 days and reach the level of 99, (9)% in the list of requests for 110 days of the search engine.
Even if we reduce the share of new requests to 0.1%, it’s still the milestone of 99, (9)% comes amazingly fast.
Source: https://habr.com/ru/post/131012/
All Articles