Link diversity has been widely used as a means to rank a site, but what role does it exactly play in it? Is this role still relevant in 2018?
I wished to make things clearer so I asked for the opinion of an expert in search engines – Ammon Johns. According to Ammon Johns, people have noticed how sites can have diverse and unique links that don’t indicate manipulation. This observation then led to today’s perception of “link diversity”. He pointed out what link diversity is: it’s all about the number of links coming from an authority site that Google believes can’t be bought or manipulated. Link diversity can be likened to peeling a sticker: grabbing a corner allows you to peel the entire sticker away.
What exactly is Link Diversity?
Link diversity is about obtaining inbound links from different kinds of pages and variety of domains. It involves obtaining a variety of anchor text in both “follow” and “no follow” links. “Follow” links pass link equity, while “No Follow” links don’t pass page equity, and so, don’t improve PageRank. (PageRank is what Google uses to determine a website’s popularity.) When a site has gathered enough diverse inbound links, its backlink patterns will be kept within a normal range and won’t appear “spammy”. Natural-looking backlink profiles help the site escape detection from the algorithms of statistical analysis.
Statistical and Modern Link Analysis
Statistical link analysis is the field in charge of studying inbound and outbound links’ characteristics. First used in 2005, statistical analysis analyzes factors such as the number of outbound links, inbound links, etc. The aforementioned factors are graphed, and sites that have these factors form a gentle curve. Those sites that fall outside of that curve are the spam sites; they cluster outside the curve and are called outliers because they fall outside the standard ranges. Modern link analysis, on the other hand, uses groups of trusted sites wherein Google takes note of link distances between authorized sites, spam sites, and the sites they link to. With modern analysis, bad links have nowhere to go – they can’t act as if they are just one of those regular, non-optimized links. Link diversity does not matter when distances between sites are calculated. Links are measured based on the quality of sites linked to it, and link diversity is simply not a factor in this.
Diversity, but not Link Diversity
Search engines will only consider diversity when it comes to relationships among sites. Search engines identify ownership information; this helps them determine links that are being manipulated. Let’s say someone owns a hundred sites. Linking patterns, whois, and hosting IP information can then be used by search engines to detect sites linked to one another. They can also determine attempts for link selling and spamming. What’s important for search engines is to remove links that are irrelevant, as well as those links used for link manipulation. At the end, only the relevant links remain to be involved in the ranking process.
Dismissing Link Diversity
There are two reasons for dismissing link diversity:
• It serves as an excuse for poor link building practices. Those who intend to use links of poor quality will just end up playing the “link diversity” card to justify their behavior.
• Search engines have already changed. Search engine algorithms change from time to time. Link diversity may have worked for some back then, but there’s no special reason to do this now.
Simply put, link diversity no longer works at this day and age. You don’t have to have links from sites of different domains – top ranking sites’ backlinks can prove that.
No research or patents are known to support the idea of link diversity being a valid ranking signal, and so, there’s no need to implement this practice. Still wish to implement link diversity? Go ahead, but you should make this a priority when discussing SEO strategies.