Emergent Properties

One of the best algorithms people have come up with for providing good search results is Google’s PageRank algorithm.  Simply put, the value of a specific page on a website is related to the value of the pages that link to the page.  So, if lots of people link to you, you get higher PageRank.  The key here is an actual, real, HTML anchor-text link.  No Javascript clicks, no banner ads, etc.  Those things don’t count.  The value of a page is related to how many REAL links there are to it.  Also important is the value of the pages that link to a specific page.  So, if a page gets linked to by a highly valued page, it gets ranked higher than a page that is linked to by a less valued site.  It’s pretty simple.  If a site is linked to a lot, it’s an authority.  If it’s an authority, and it links to you, well, you’re good too.  It’s fairly common knowledge that PageRank isn’t the ONLY thing that Google looks at, but, regardless of what anyone says, it is important.  Google conceived of the system, and they provide pretty good search results.  Everything’s great in the world.  Except… when it’s not.

Looking at web history, webmasters figured out the importance of links, and they did what they could to get them.  Link exchanges became common, and they still are.  Link spamming, especially on blogs, became common.  Google had to react, to make sure that their algorithms still worked.  If certain people are unscrupulously gathering links and are performing better in search results, Google ceases to become an independent provider of relevant search results.  If this happens too much, people stop using Google.  So, the point is that people have been trying to game Google, and Google has been reacting.  However, for the most part, Google’s been reacting to things like link spamming, blog spamming, and the like.  But now, things are changing.

Text Link Ads is a service that lets webmasters BUY links.  This is different.  Instead of swapping links or having a site’s content be editorially dictated, there’s now a market for buying links on sites with good PageRank.  Webmasters like it because it gives them another stream of revenue, people doing SEO like it because it gives them an easy way to build links.  Text Link Ads says that it’s not really about PageRank.  I disagree.  If it weren’t about PR, then why not build something resembling a banner ad system?  Why not value links based on clicks or impressions?  Because you want real anchor text links that translate into transferred PageRank, that’s why. (Note: I’m not saying there are not other reasons for buying TLA’s, such as traffic.  But If PR weren’t a large part of this equation, why not just have a more traditional ad system?)

Is this bad?  My guess would be that Google probably thinks so.  Google wants their search results to be good, and anything that potentially hurts that is a “bad” thing.  However, the case can be made that this actually isn’t so bad.  For one, another revenue source brings more revenue to websites, promoting better content.  Also, TLA has editorial approval of links (and sites displaying links), and for the most part, the links actually conform somewhat to the niche they’re being displayed in.  Here’s an analogy:  You can pay Yahoo to be in their commercial directory, and Google counts this as ok.  They even suggest it.  The idea is that if you pay Yahoo to list you, and they do list you, it’s approval that you’re good.  You’ve spent the money to prove that you’re not a fly-by-night scam or spam operation, and Yahoo has independently confirmed.  Google likes this.  Take this to the next level.  Would Google approve of other directories doing the same?  Why?  I think the same rules apply. Take it another step — to where Text Link Ads is.  They’re still like a directory in that they place links in niches.  There’s still editorial control, most of the links are in-genre, the real money exchanged is a barrier to uberspam.  It’s just … decentralized.

I think TLA itself isn’t enough really hurt Google.  However, there are networks out there that take this idea another step towards spamminess.  Here’s what they do:  You give up links on your sites for links on other sites.  It’s just like TLA, but you get “paid” in links which you can only “spend” on other links.  These sites generally skim off the top and actually sell some of their inventory.  Also, they’re fairly sophisticated about how much “currency” you get for displaying links on your site.  Here’s why they’re bad:  There’s much less of a barrier to spam and lack of editorial control.  Once systems like these become automated enough, they’ll degenerate into spam.  On the other hand, maybe they won’t.  Maybe there’s enough money to be made and enough focus on keeping everything legitimate that they won’t really hurt google that much.

However, can it be said that something that gives some sites an advantage over others in the SERPS is not basically bad for Google?  In the end, don’t these sites exist to game Google?  Isn’t anything that games Google bad for Google?

Time will tell whether these networks thrive or whether Google figures out how to do some fancy graph analysis that penalizes network participants, thus killing the networks.

What do you think?  Is Google screwed by such practices? Will they evolve? Is a solution possible?  Does it require more computing power than is possible (even for the Googleplex) to amass?

Time will tell.

technorati tags:text+link+ads, google, text+link+ads+google