Quick Update: Make no mistake, building PBNs is still a viable SEO tactic. After Google started to crack down on PBNs, we decided to focus more on businesses models that don’t only focus on SEO. However if your goal is to learn more about PBNs, the information below is still very useful.
What
The goal of the experiment is to identify the typical link quality pattern for authority sites.
Why
The reason for the interest in this is to determine if a money site using a PBN has a different link quality pattern than a naturally-linked site. The idea would be to compare the “natural” profile to money sites using a PBN. If there is a significant difference, we can recommend how to build a more natural link profile so there is no footprint that Google can detect.
Of particular interest, is the quality of links as a site first launches since this is the time when a PBN will be providing most of the backlinks. It is expected that as time goes on, natural links will grow.
How
A sample of authoritative websites needs to be identified so the natural DA profile can be created. The authoritative sites will need to contain informational content, much like the money sites to be analyzed. In addition, the number of pages of the natural sites should be similar to the money sites. Other matching factors may be added beyond the two just described to have more of an apples to apples comparison.
After the sites have been identified, Referring Domain information from Ahrefs api will need to be collected. From this data set, we can identify backlinks occurring during the first six months. The DA for each referring domain can be obtained through the Moz api.
The simplest analysis will be to compare the DA distribution of the natural profile compared to the money sites.
A more sophisticated analysis may be to see if a predictive model can identify Money sites vs Natural sites. Model parameters would include DA, linking frequency, Alexa rank, backlinks count from one domain, etc.
Who
The most challenging part will be to identify natural sites to compare to money sites. The help of the NoHatDigital team will be needed to suggest comparative sites and to validate these sites are naturally link built.
When
The experiment should be complete by the end of September
Update: 8/18/14
Dofollow backlinks to over 100 money sites were identified using Ahrefs. Links seen during the first six months for which Ahrefs has a record were identified. The DA of these referring domains were then used in the analysis.
DA was categorized in 10 point increments (0-10, 11-20, etc).
As you can see, only 1 percent of referring domains were DA 1-10. This seems somewhat unnatural although the pattern cannot be confirmed until we have a comparative set of sites. We will also be evaluating the pattern for DA 40 and above backlinks as 32% of the referring domains fell in this range. Intuitively this seems a bit high but let’s see what the science says.
| DA | Money Sites With Backlinks | |
| Count | % | |
| 0-10 | 28 | 0.01 |
| 11-20 | 480 | 0.24 |
| 21-30 | 486 | 0.24 |
| 31-40 | 365 | 0.18 |
| 41-50 | 193 | 0.10 |
| 51-60 | 172 | 0.09 |
| 61-70 | 106 | 0.05 |
| 71-80 | 58 | 0.03 |
| 81-90 | 44 | 0.02 |
| 91-100 | 76 | 0.04 |
The next step is to identify a comparative set of sites and perform the same analysis.