By Sage Lewis, The ClickZ Network, Sep 13, 2007
Since SES San Jose, I’ve been thinking a lot about this whole “Are paid links evil?” debate. Google says, “Don’t buy links to increase link popularity.” Web site owners say they have little choice because no one will link to them otherwise. Most mainstream SEOs agree aggressive linkers don’t deserve a halo over their heads. There is no question the search engines’ combat against aggressive linkers has evolved.
In fact, it shook me a bit when I realized how easy it is for a search engine such as Google to detect sites using paid links or boilerplate linking tactics.
So, I thought it might be interesting to put myself in Google’s shoes and imagine myself as a semi-human search engine trying to solve the paid links problem. What if I were tired of all these aggressive link-building sites manipulating my algorithm? What would I do if they were wreaking havoc with my SERPs (search engine results pages)? What if I wanted to strongly penalize the aggressive link-building offenders?
Visualize me as “I, Google,” a cyborg search engine imbued with superhuman powers to deal with aggressive link builders and paid-link buyers.
“I, Google”
Here is my recipe for addressing the paid-links dilemma. First, set a benchmark for link relevance. Audit potential danger zones such as the personal injury, pharmaceutical, and gambling categories. Rate the relevance factor for highly-competitive, highly-searched phrases in these categories.
Create an internal focus group to rank the relevance of the top ten results for keywords such as “personal injury,” “Viagra,” and “poker.” On a scale of 1 to 5, rank relevance with 1 being “completely irrelevant” to 5 being “completely relevant.” The results would be recorded and tested against the same focus group after applying an aggressive linking filter.
Create a tag called “LinkingTarget” for sites that meet the criteria identifying them as aggressive linkers. This filter would then be applied to any keyword where more than 100 sites compete for organic search rankings. The filter would also override the standard Google ranking algorithm.
However, don’t kick out the offending sites. “I, Google” doesn’t ban any site that may be innocent until proven guilty. Besides, aggressive link builders are often loud and might just call more attention to themselves. Instead, put a ceiling on their ability to rank highly. Keep the offending sites at position 11 or lower in the results for keywords with over 100 competing Web sites. Only allow the offenders to come up for very niche phrases such as brand name or Web site name.
Tag and Monitor Offenders
In our initial round of testing, a site would be tagged with “LinkingTarget” if there is a sudden increase in links or crosslinks when none existed before from the following:
- Sudden increase in links from sites with PageRank 4 and above where none existed before.
- Sudden increase in links from blogs where none existed before.
- Sudden increase in links from second-tier directories where none existed before.
- Sudden increase in links from non-related sites.
- Sitewide links from outside sites.
- Crosslinking with sites owned by the same person.
Check for sites that meet the above criteria within the last 365-day period. If all the above criteria were met within that rolling time frame, apply a “LinkingTarget” tag and attach the filter to the site.
Six months after the tag is applied, ask the “I, Google” focus group to rank the relevance of the original phrases. Compare relevancy scores from the previous test to see if relevancy increased across multiple phrases and industries. Check to see if the results seem to be random.
After a test period of six months, sites tagged with “LinkingTarget” would be monitored every 90 days for changes within the aggressive link-building criteria. If a decrease were found in two or more criteria, the “LinkingTarget” tag would be removed.
Link Crime to Punishment
The “I, Google” process described above would only penalize the most blatant link builders. Only sites meeting all criteria are likely to be guilty of building links strictly for increased link popularity. Plus, this system would be hard to game. The chances of a competitor initiating an attack campaign would be slim because the filter is highly confidential. Companies engaged in building links solely for link popularity are likely using most if not all of the proscribed link building tactics. Many sites trying to manipulate link popularity would be affected by this filter. Based on the success of the initial “I, Google” campaign, criteria can be adjusted to improve relevancy.
Eventually, the “aggressive linker” community would be demoralized. With steady pressure, link building solely for increasing link popularity would become an SEO practice marked for extinction.
