Is Google Using “the Temporal Pattern” to Detect Paid Links?

Share This

I just finished reading a post called, 6 Links that Look Like Paid Ones but Really Aren’t. It brought up some good points about how innocent sites might inaccurately get hit with penalties for accidentally looking like they sell text links.

“This post was written as an illustration to show how difficult it is to detect a paid link, whether it’s algorithmically or manually.”

It’s a great post, but I wanted to add an additional way that Google might be able to detect paid links.

Here are the original six methods from wiep.net:

  1. The Unrelated sidebar links
  2. The Related sidebar links
  3. The Sponsored links
  4. The Reviews
  5. The Footer Links
  6. The new link in the old post

One additional pattern of text link buying is the temporal pattern. Text links are often bought in month-long blocks of time. If sitewide links to external sites continually appear and disappear on a site in 1-month blocks, it is a possible indication of paid links.

I don’t think that Google uses any one factor in their paid-link detection algorithm. It would be a combination of factors. Maybe month-long blocks of time combined with a group of 5 or more sitewide, external links and an affiliate link to text-link-ads.com?

For detecting reviews, Google might look for the presence of reviews as well as the presence of links to paid review services. Or they might look at a certain number of off-topic review posts combined with the presence of links to paid review services.

Google has so much data on the Web that they could datamine their index to develop “Web site profiles” of sites that sell links. For example, they might be able to determine “if a Web site has factors x, y, and z all at once, there is a 99% chance that it is selling links”.

I don’t buy links anymore, but I’ve (reluctantly) done it in the past so I find the debate interesting. In certain industries it might be hard to compete with people who buy links unless you do the same. It’s difficult to compete with old sites that got their authority through link exchanges and link buying, and that are now old enough to keep their authority. The sites that have paid links to your competitors aren’t going to link to you unless you pay them.

Or—you could just buy AdWords

Google cannot stop the paid link industry. It will just become more sophisticated. People will learn to mask their Web site “profiles”. The prohibition couldn’t stop alcohol use, it just made it more expensive and secretive…

2 Comments

  1. Posted October 11, 2007 at 12:59 am | Permalink

    Hmm, good point, I didn’t even think about it this way. I agree with you that a section of fluctuating links should be quite easy to spot. However, this can also be a natural thing, for example a block of “websites of the month”.
    You could take this from the other side as well; it’s quite easy to spot if the majority of a website’s incoming links are temporary links.

    I guess changes like this move Link Building more towards Link Profile Management.

  2. Posted October 11, 2007 at 4:15 am | Permalink

    Thanks for stopping by…

    If Google uses this kind of technique, a block called “Web sites of the month” might get a site in trouble.

    I think Google might say to themselves, “we have found a technique that is 98% accurate — the 2% of innocent people that accidentally get hurt are just collateral damage. We can’t build a perfect system… and they can just go buy AdWords until their problems are sorted out in the PPC-laden ‘Google Webmaster Help’ Google Group”.

One Trackback

  1. […] I wrote a post about how Google could be using the temporal patterns of paid text link buying to detect paid links. I wanted to add more examples of how Google might algorithmically detect paid […]

Post a Comment

Your email is never published nor shared. Required fields are marked *

*
*
Close
E-mail It