Google has maintained their market leader position around the world through constant innovation aimed at ensuring the best user experience across their primary search product. Therefore Google reward sites that offers the best, most relevant, authoritative and most natural user experience to their visitors through improved search rankings.
Traditional SEO practice around anchor text use for backlinks has been to focus on increasing the sheer number of backlinks to their sites by capitalising on as many mentions, or citations, of their site/brand as possible. This has historically resulted in text content on many sites becoming unnaturally ‘stuffed’ full of anchor text hyperlinks and thus had detrimentally affected the overall user experience of visitors reading such content. Google have countered this trend through the release of the Panda algorithm by analysing the signal-to-noise ratio of a sites’ backlinks – that is, the number of site/brand mentions in the text versus the number of backlinks to that site.
The engineering concept of ‘signal-to-noise ratio’ has been traditionally defined as the ratio between the strength of the signal carrying wanted information compared to unwanted information signals in the background. This translates quite easily to the world of SEO in the sense that the first mention of a site or brand should be linked to the respective site using anchor text as this is useful to the reader should they wish to conduct further research. However future mentions of the site or brand should not necessarily be linked to the respective site as this can become distracting to the overall reading experience, hence becoming unwanted noise.
The Google algorithms detect this ratio by comparing the overall number of backlinks to the number of unique domains these backlinks originate from. Through this analysis Google can determine which sites require further scrutiny should they have a high number of overall backlinks originating from a very small pool of domains, which can be indicative of such anchor text spamming which can reflect a poor user experience as explained above.