I would love to hear about some of the specifics of your test. I would like to compare your results with some of my own data.
To be absolutely clear, I did not set up a test where I created a bunch of spammy backlinks. My data comes from what I perceived to be attacked by competitors that waged a series of spamming campaigns targeting a number of websites that I operated as well as sites owned by some of my clients a couple of years ago.
I first noticed this when I spotted some traffic coming from unusual keyword searches. On closer inspection, I discovered that tens of thousands of very spammy backlinks had been placed on several thousand websites, all of the dubious characters. (we’re talking porn, viagra, warez, etc. and many were hacked sites and loaded with virus-infected files). First, on just one website then every couple of weeks a new website would be hit.
These spam attacks took place over a period of about 9 months and then the backlinks started to gradually disappear. I don’t think there would be anyway a search engine could distinguish an apparent attack like this from a webmaster’s own spamming activity.
These attacks targeted websites that were in competitive markets and they each had pages that ranked #1 in SERP for a number of highly competitive keywords within their niche. We also had some fairly expensive PPC campaigns targeting the same keywords. The business we were generating from these terms was in the 6-7 figure range, so I was a little apprehensive about how this might affect our campaigns.
I’m happy to report that our SERP positions never wavered during any of these apparent attacks and the only downside was the embarrassing backlink profile that could be seen by anyone who knew where to look.
I would be very interested to learn the differences between your experience and my own, seems like there would be some valuable insight in that knowledge.