Four extremely intelligent Californians, Zoltan Gyongyi, Pavel Berkhin, Hector Garcia-Molina, and Jan Pedersen, released a whitepaper discussing a possible method to identify links from spam domains. Their method sheds light on some very interesting ideas on how Google might judge the quality of inbound links.
Good links Vs Bad links
Since the Jagger update, Google seems to be emphasizing link relevance and link quality a lot more. While relevance is pretty straight forward to understand, quality of an inbound link may be a foreign subject to those not involved in search engine marketing as part of their daily lives. So let’s first discuss good links versus bad links. Not all links are equal. Links from popular sites carry more weight than links from virtually unknown sites.
It is a well-known fact that Google’s initial success lay in their ability to find websites and rank them based on links from other websites. As Google’s popularity grew, some webmasters tried to outsmart the Google algorithm by falsely inflating the importance of their website with unnatural inbound link development. Links that are developed in such a way, with the mere intention of fooling Google’s algorithm to gain higher rankings are considered to be bad links.
What is Spam Mass?
The research team from California devised a term – spam mass – to denote the ratio of good links to bad links for any website. Simplisticly speaking, if 60% of the inbound links to a website originate from spam domains – domains identified to be built solely for the purpose of artifically inflating a website’s popularity – then the spam mass of the site would be 60%
Ranking is not that simple, though, and variables such as PageRank, relevance, age etc creep in to complicate such a calculation. However, the idea bears a lot of potential from which further ideas can be developed.
Applying Spam Mass to Search Engine Marketing
– Why reciprocal link exchanges can pull down your site rank
When optimising a site for better search engine ranking, most webmasters and search engine optimisers will actively seek inbound links from third party websites to their site. A very simple way to get these inbound links, in the past, was to exchange links with other site owners. The problem arises when most of your inbound links are from sites whose own PageRank is also influenced by link exchanges.
Google is getting better at recognizing when search engine optimisation relies on spammy link development to improve rank. The principles Google uses to do so, are likely to be based on concepts similar to spam mass. For example the Google system can differentiate links from good domains and spam domains.
If the PageRank of your site is mostly due to links from spam sites, Google will accordingly devalue your PageRank, discount any effect from inbound links from the spam domains and, in extreme cases, label your domain as a spam domain too! All of this can lead to a reduction in your site’s rank on the search results.
How to avoid Google penalties for spam links
The easy answer to this is always “build your site for people, not for search engines”. That doesn’t mean you do not carry out the necessary search engine optimisation. It just means that you should not indulge in unnatural practices to artificially influence rank. Ideally, let your site and rank grow as if it were a natural progression.
- Do not indulge in reciprocal link exchanges with spam type domains
- Stay away from link farms
- Exchange links with your suppliers and customers
- Get links from trade bodies, governing organisations
- Exchange links with competitors, other related sites
- Provide links to relevant authority sites
- Strive to obtain as many high PageRank relevant inbound links as possible
Read the full technical report