One thing we need to be aware of in all of this is that Google’s mission is to return the best Web page for each result in order of most relevance. So, even though from a mountain-top perspective it seems like the farmer update affected a lot of user-submitted content-sites, the fact is that it really only affected a lot of user-submitted content.

In order for us to be successful in spinning, posting, commenting, linking, pinging, mentioning…anything we do to build links and reputation, it needs to fit within Google’s own expectations for a given web page that is a subset of another group.  Such as, an online store that’s just starting can easily grow by 50 pages a day for a while.  A debt settlement site that’s just starting out, maybe that can only grow a page a day for a while and then increase a bit at a time.

Here’s where the site matrix becomes important.  A site that has 4 original content articles probably won’t get 50 links a week.  But, a site that has original content articles that’s a 1000 pages deep….getting 50 links a *day* is easy for Google to accept as natural.  So that just means as long as you don’t push the link-building needle to red for a site that doesn’t fit the criteria for a site that should receive that many links, you won’t have a problem.  Push the needle too hard and you can expect your multiplier to drop; drop enough and you’ll experience the sandboxed feeling until.  But, pages are always being ranked so you’ll “magically” reappear one day when it was just a matter of you continuing to build out your site in a consistent manner.

Additionally, the farmer update was probably a matter of improving the duplicate content filter to include variables I am only able to identify conceptually.

Here’s a scenario, (if Google was a person…no commentary on whether or not that will someday happen)…Google goes out and finds all these links.  Google will base it’s ranking decision based on certain criteria, (we’ll call this ‘step 1′).  When was the domain registered? When was the last DNS change? When did certain pages get indexed.  How did the site grow?  Was it slow and steady?  Was it fast and steady?  Was it sporadic and vacillating between non-existent and fast?

Apply that same concept to links.  How many links? How often?  Always the same link text? Little variations? Unpredictable?

The way Google handles step 2 is based largely on what he/she discovers in Step 1.  If everything measured in step 1 fits within an acceptable or expected pattern for an organically growing Web site then the result will act as a multiplier for each page on the site.  If the multiplier is .75 then Google suspects the site is bit spammy but not completely overdone.  If the multiplier is 1.25 then you’ve done an above average job of organically growing your Web site or at least making it appear that way.

In Step 2 Google looks at the content surrounding the links and tries to determine if this content is valuable.  Essentially, the farmer update appears to be some sort of improvement to the duplicate content filter.  Is this spun content?  How often do certain words or phrases appear on pages that host your links?  How similar is the structure of the articles?  How varied is the length?  Now, where are these?  A page hosted on a site with a low multiplier will send low-value links.  Plus, if you have spun content with little variation in the structure of the article, regardless of the content; it still could look a little spammy.  Put the evaluation in step 1 matched up with step 2 and you can see that the farmer update will only hurt bad content but directly only impact individual Web pages and the weight the links on those pages are given.

This is an ongoing process; the ranking of pages is always happening.  Newly discovered links are instantaneously scrutinized.  Google wants to know their link-family-history and connections.  When a new link is introduced 6 levels up, it will affect all links in the chain.

How do we solve this if we’re just trying to SEO spam?  Well…we have to get better at diversifying the structures of our spyntax for one.  Next, we need to sometimes link and sometimes only mention.  When we link, we need to link the way people really link and even sometimes include our competitors in the mix, (choose somebody big like Bank of America).  We need to have a myriad of links that we can randomly draw from so that we’re passively promoting aggressively.  (meaning we’re promoting all the time but just don’t know specifically what that is…just that it has something to do whatever the subject is)  We need to build articles within articles within articles that all have such a varied structure that any article spin will generate anywhere from 200 words to 1500 words but always look completely unique as far as structure, link(or mention), anchor-text(if any), and the links themselves.  Every spin needs to read well as if a real human dripped blood, sweat, and tears into that article.  A real human or group of humans did…for sure; but just not a million different times.

I view all spyntaxed articles as living documents without a definite end-date.  Like a Web site, spyntaxed articles will continually be improved upon and will grow more indistinguishable as spun or rewritten content, (guess the second half of that isn’t a characteristic of a Web site).

Alright, there’s probably a lot I missed; but that’s the basic idea.  I have to get back to work.

Thanks,

David