I have been doing a fair amount of back-link research on my competitors. I have noticed that those with large back-link profiles have set up either a cdn utilizing a separate root domain with multiple subdomains i.e. images, pdf's, video's. This separate root domain accounts for nearly 1/10 (100,000 links) of the overall back link profile.
What is baffling me are the back-links coming from the cdn to the main domain. I cannot find any pages or resources that link back to the main domain from the CDN. Obviously the main domain points to the cdn for the references to images, pdf's, and video's. Any ideas on how they could gain back-links from this strategy?
The CDN is gaining link juice from all of the follow links from the main to the cdn but I cannot understand the reverse and how it accounts for 100,000 back links. I am intrigued by the strategy because it seems to work but how is this not direct spam?
I have checked all subdomains and root domains for a robots.txt file but have not found one.
We have a cdn network and have a robots.txt to nofollow everything but I am now thinking that this should be changed to nofollow just pages but all other assets should be indexed, anyone have prior experience or reccommendations?