Payday Loan Affiliate Network
Home Sign Up for a Payday Loan Payday Loan Blog
Apply for your payday loan today and get approved for up to $1000.

How Our Program Works
Frequently Asked Questions

Payday Loans
Affiliate Marketing

Add This Blog to your Technorati Favorites
Subscribe to our RSS Feed

Payday Loan Blog - How To Outthink A Room of PhDs

 Friday, 13 April 2007

I struggled for a bit on the title because the point of this post isn't to come of as pompous or boastful; rather, it is to instruct on a general way of thinking. A while back, I drew a rather simplistic view of what a search engine is, comparing it to a single data table with a few columns (scoring variables) and a few rows (some URLs). Based on that simple concept, hopefully at least a few people began to realize that the goal of what search professionals do is to try to anticipate what variables will be most heavily weighted, and build sites around those variables. In time that variable mix may be quality content that is appropriately viewed upon by its true peers as important relative to other documents, but that mix just isn't there yet.

So, if the variable mix isn't to the point where quality content that is viewed as quality by its true peers, what is the goal of a search engine? To get there.

How do they get there? Incremental changes usually. We've all seen the changes as search engines evolve, chasing that relevancy perfection that provides the optimal blend of paid advertising and pure unadulterated content…from content to votes, from votes from really authority sites, to votes from really authority sites, etc; the engines do and will continue to evolve.

So, how do you outthink a group of PhDs? By not trying to outthink them.  Are you confused? It is slightly counterintuitive, but you'll see what I mean.

With all the A/B testing that goes on, tweaking filters, employing re-ranking tactics, constantly shuffling and reintroducing data, there can be a whole lot of things to track in your attempt to stay on top of the game. Try not to focus too much on that for the time being though - instead, I want you to develop a list of phrases to track on a fairly regular basis, completely independent of each other. Track historically spammy phrases in the male enhancement pharmaceutical arena to dog biscuits, civil war action figures, to the bicentennial - track a lot of competitive and non-competitive phrases because I want you to be able to see what happens when changes occur. Every week, run through your list of phrases and just look for a few things:
1. Top 10 players, their top 10 incoming links according to Yahoo, and total # links
2. Those that fell out of the Top 10, their top 10 incoming links according to Yahoo, and total # links
3. On each phrase, is the current mix more relevant or less relevant [score 1 for yes, and 0 for no]

Whenever you want to know what the engines currently like, just compare the two sets of data…really, that simple usually. When I say compare, that becomes a bit more nebulous because what I personally look for isn't always the same…sometimes I may notice a keyword phrase within the URL being given more weight, sometimes it might be keyword density, age of site, co-occurrence of related phrases, etc. Usually though, in our current environment, it has been links. How many, how powerful, how relevant? If the score is higher than last week, and you were being honest, expect the changes to stick; if it was lower…a lot lower…you can probably go back to doing something else.

Now, a room full of PhDs would probably assume that some people are going to try to reverse engineer their algorithm and filter sets by looking at a large grouping of unrelated data; thus, what you just did isn't all that earth shattering to them, and I have a hunch that they wouldn't really care, given that for most search professionals, the next logical step is to provide the engines with that they are trying to reach: relevancy.  Yes, holes will be exploited, but so long as their A/B testing shows a measurable increase in overall relevancy from their own grouping of unrelated keyword phrases, that change is likely to stick.

If you can see a certain variable weighting change that would improve relevancy, plan for it occurring; build a site just for that purpose if you have to.  For instance, I think social media marketing will get filtered down a notch in the upcoming year, due to its rampant use, either by taking into account how quickly the links are added (already there, but just tweaking that) and by taking into account more differences between root domain theme / page theme, root domain authority / page authority, inlink theme. Who knows though, the A/B testing may show that the end results of social media marketing actually produce more relevant results.

Have a pleasant weekend.


Friday, 13 April 2007 08:15:44 (US Mountain Standard Time, UTC-07:00)  #     

Call Toll-Free Now! 855-371-8009