Search analytics is the analysis and aggregation of search engine statistics for use in search engine marketing (SEM) and search engine optimization (SEO). In other words, search analytics helps website owners understand and improve their performance on search engines. Search analytics includes search volume trends and analysis, reverse searching (entering websites to see their keywords), keyword monitoring, search result and advertisement history, advertisement spending statistics, website comparisons, affiliate marketing statistics,multivariate ad testing, et al.
Search analytics data can be collected in several ways. Search engines provide access to their own data with services such as Google Trends and Google Insights. Third party services must collect their data from ISP's, phoning home software, or from scraping search engines. Getting traffic statistics from ISP's and phone homes provides for broader reporting of web traffic in addition to search analytics. Services that perform keyword monitoring only scrape a limited set of search results depending on their clients needs. Services providing reverse search however, must scrape a large set of keywords from the search engines, usually in the millions, to find the keywords that everyone is using.
Since search results, especially advertisements, differ depending on where you are searching from, data collection methods have to account for geographic location. Keyword monitors do this more easily since they typically know what location their client is targeting. However, to get an exhaustive reverse search, several locations need to be scraped for the same keyword.
Search analytics accuracy depends on service being used, data collection method, and data freshness. Google releases its own data, but only in an aggregated way and often without assigning absolute values such as number of visitors to its graphsISP logs and phone home methods are accurate for the population they sample, so sample size and demographics must be adequate to accurately represent the larger population. Scraping results can be highly accurate, especially when looking at the non-paid, organic search results. Paid results, from Google Adwords for example, are often different for the same search depending on the time, geographic location, and history of searches from a particular computer. This means that scraping advertisers can be hit or miss.
Taking a look at Google Insights to gauge the popularity of these services shows that compared to searches for the term Adwords (Google's popular search ad system), use of search analytics services is still very low, around 1-25% as of Oct. 2009. This could point to a large opportunity for the users and makers of search analytics given that services have existed since 2004 with several new services being started since.