Top Three Query Returns Increase Value, As Exact Term Positions Decrease Options

[iamlost]

The seo(who-does-not-sleep)bythesea has an extensive look at a Yahoo patent giving insight into an increasing SE behaviour:
How Search Engines May Substitute Other Search Terms for Yours

What is critically important to those optimising sites for SEs (by definition a much larger pool than those offering SEO services) is the drastic decrease in available real estate. Not only are you now (and, imo, increasingly in the future) competing on exact query term but also on related query terms.

Let’s say that a large number of people who search for the term intellectual property then go on to search for the term patent attorney with their very next search, or within the same search session.

The search engine log files would uncover that such an association exists, and the search engine might explore how common it is for searchers to search for that second phrase. If it happens frequently enough, the search engine may start suggesting patent attorney as a suggested search to searchers along with a display of search results for the term intellectual property.

This means that if you currently hold one of the top three positions for a term you can work returned associated term(s) as well for increased presence. Unfortunately, it also means that if your site shows at lower page positions it may be bumped off the first page to the second.

When you add in the also increasing tendency to split the results page to offer possible alternative meanings of query terms the first page SERP property available is fast shrinking.

A few somewhat rhetorical questions:
* What is the value increase of a top three term placement when the exact query result drops from 10 to <5 returns?
* What is the value increase of 'long tail' content when associated query terms are auto-included?
* What is the value boost to a query term SERP by highly ranked onsite associated terms?
* What is the value boost to associated long(er) tail terms of increased deep linking?

[projectphp]

* What is the value increase of 'long tail' content when associated query terms are auto-included?

IMHO, the long tail is the LEAST likely area to have this stuff done, for the very reason that a long query is specific. Lets say you substituted laptop for lap top, but soemone was searching fo a brand name and model that incldued the word, a substitution wouldn't make a lot of sense, nor would it likely improve results.

Add to that the fact that this sort of thing adds a layer of complexity (and therefore slows things down), and for longer searches, it is very unlikely to be the norm, at least initially.

[iamlost]

Perhaps we have a different understanding of 'long tail'.

Mine:
The web market, given it's minimised distribution and inventory costs, maximised choice and traffic, conversion pattern defaults to a power law distribution curve of 28% popular/hits/head and 72% unpopular/non-hits/long tail (offline curve is the more commonly known 20%/80%).

Thus long tail is not only >3 query terms/modifiers but singular terms outside the most popular 28%. What exactly is head or tail varies by site and niche; in this instance it is whatever a particular SE happens to say it is.

Remembering that the area (or total traffic/conversion) under the tail can exceed that of the head it is statistically probable (and has been seen in reality) that a 'long tail' term queried frequently as a second/third/fourth choice after an initial 'head' term query would be shown to influence actual returns.

What this patent suggests is using stored historical data to save some subsequent searches.

I have some reservations (especially about Google) as subsequent searches are actually good for the search ad bottom line. However, given the potential of broadening the ad match criteria on a single page it might well cost out better.

[bragadocchio]

Thanks, Iamlost.

I've been hoping to see a patent filing from one of the major search engines for a while that went into more details on the kinds of assumptions that a search engine might make when it decides to offer alternative query suggestions or blend results from additional query terms.

While the patent application discusses Yahoo, we can see something like this working in Google. For instance, look at the search results for

icecream sundae

Above the search results, we are asked:

Did you mean: ice cream sundae

The set of search results for each version share some pages, but they do differ. It does look like the results from query with the two word version of "ice cream" are blended into the results for the query with the one word version of "icecream"

What is critically important to those optimising sites for SEs (by definition a much larger pool than those offering SEO services) is the drastic decrease in available real estate. Not only are you now (and, imo, increasingly in the future) competing on exact query term but also on related query terms.

There's an escalation of impact when a search engine goes through a process like this:

First - the search engine doesn't see any similar or related search queries from the query logs or co-ocurrence data or dictionary like Wordnet , so your search results focus primarily upon the query terms used, and you aren't offered alternative query suggestions at the top of the results

Second - the search engine does identify some potential alternative query suggestions, and provides links to them at the top of the search results, but doesn't blend results from the alternative into the search results.

Third - the search engine identifies one or more alternative query suggestions, and blends results from that alternative query into the search results for the original query term.

I'm not so sure that whether the original term searched for is a long tail term is important - what might be in many cases is whether or not people tend to reformulate their query after performing a search for the first query. For many long tail searches, that is a possibility since people may find that the results for a long tail search may not provide enough results, and they may shorten or change their query terms.