IR researchers on SEO

Xan is one of the more interesting posters at the Search Engine Watch forums. He apparently comes from the Information Retrival side of search - the research rather than webmastering side.

What I find most interesting is that he is apprently quite fascinated by SEO, and goes so far as to consider it as possibly even useful, rather than condemn it outright.

He posted a set of observations in his blog concerning SEO forums, but falls into a trap of thinking all SEO is monetised:

SEO is undertaken in order to obtain monetary gratification of some kind.
Computer science is an academic and scientific effort, exploited for money.

SEO isn’t necessarily about monetisation - its fundamental basics apply to being able to associate websites with keyword areas.

In an ideal world, Mr Fred’s site about pink elephants would be associated for relevancy with the search phrase “pink elephants”. In the real world - and a fact that many sites fall victim of - is that he might only have the words “Pink elephant” in a title graphic. And the main index page simply be titled “Mr Fred’s site”. And the pink elephants that his site is about might only appear in images of pink elephants, with basic commentary and information possibly accompanying covering only location and timing.

In short, Mr Fred’s site about “pink elephants” might never even contain a single indexable word about “pink elephants”. It could be missing from his text, and his graphics could easily be missing descriptive alt text, if they are even present at all.

To any passing search engine, this site is simply about Mr Fred and photo locations.

This is the world that SEO grew out of.

SEO said - “Hey, if we use some descriptive accessibility features, we could make search engines realise your site is about pink elephants!”

So the word “pink elephants” somewhere entered the title attribute; meta tags were built to describe that this site is actually about pink elephants; h1 and h2 tags were used to designate not only that this site is about pink elephants, but also that it covered a specific topic area about pink elephants (”Photos of Pink Elephants”); the word was specifically used in the text, just to help; alt tags were used to describe that each photo was about a pink elephant; and even the footer copyright used anchor text linking to “Mr Fred’s pink elephant site”. Some directory listings that had cateogries on pink elephants were also submitted to as an additionally useful tactic.

That’s the basics of SEO.

SEO isn’t first about monetisation - it’s simply a form of information accessibility. It’s a tool. And like many tools, it can be used in different ways. And that includes being profited from.

I guess I wasn’t the only SEO to learn the trade in a non-commercial environment, working on reference sites first and analysing the results.

The commercial aspect simply comes from the fact that the internet has commercialised, and where there is commercial competition, people will seek methods to move ahead of competitors.

The first hurdle in terms of search traffic, is simply in ensuring that your site is associated for the very keywords it should be associated for.

My Fred might think it unfair that he should need to use SEO techniques to associate his site for “pink elephants”. In business, it simply represents a reality that needs to be dealt with.

Depending on the fierceness of competition for the actual keyword range, this could well dictate what forms of SEO are used - cloaking, doorways, use of redirects etc, are other more advanced and extreme methods of describing content of a site page to a search engine spider.

Most of the time now, SEO’s I know are focussed on using that tool only for covering relevant keyword areas. In a commercial environment, search traffic has to convert, otherwise it is simply wasting server resources, and costing the site, rather than generating profit. And the better targeted, the better focused the traffic, and the higher the conversion rates and better revenues.

IR researchers will not like SEO because we provide an unquantifiable variable into their equations - it’s hard to reduce human ingenuinity to a mere string of numbers you can insert into an equation.

And while IR researchers may not like having to deal with SEO “interference”, if they can find someway to work with the reality that it is here and that SEO can actally provide a useful function, then perhaps IR may be able to find ways to exploit SEO for its own benefit.

Xan is like someone looking through the looking glass at a world a little bewildering and hard to fathom. Many SEO’s will look back through the same glass to IR and feel the same.

Ultimately, we can both have a shared goal - we simply want to see the best relevant content ranked highest on natural search. The IR researchers need to rank the best pages - the SEO’s will show them where they are. Somewhere in between that, there’s a balance that benefits everybody.