fantomNews™ — the ultimate know

(rt) Again, it is Mike Grehan’s e-marketing-news ezine that presents us with some refreshingly iconoclastic stuff regarding SEO tactics in its March 2005 issue: In “The Keyword Density of Non-Sense” E. Garcia takes a gander at (and trashes) some popular misconceptions regarding the much flaunted SEO technical term keyword density and the many erroneous concepts underlying it.

This is an interesting read, albeit non-propellerheads may perhaps find it a mite too technical for their taste. Nevertheless it’s worth making the effort of digging through his (highly lucid and well written) explanations and examples, if only to gain some more clarity of what KD is really all about.

“Really all about” – well, nothing’s perfect in meatspace and neither is this summary for dummies of what applied linguistics, lexicography and information retrieval (IR) technology have turned into a veritable science in its own right. Sophisticated mathematico-linguistic theories may all look very nice on paper but that by itself doesn’t imply that they constitute a realistic representation of the world out there: an experience every seasoned SEO / SEM agent has to come to terms with. I’m not sure that thinking Linearization, Tokenization, Filtration etc. as he suggests will actually prove to be of great help in Applied SEO without some more tangible, hands-on advice beyond this rather generalized statement:

I recommend SEOs to start incorporating lexicographical/word pattern techniques, linearization strategies and local context analysis (LCA) into their optimization mix.

While we would certainly agree with his basic observation that a merely mechanistic, quantitative approach to KD, keyword stuffing being a case in point, can actually do more harm than good in SEO if it ignores factors such as semantic contextualization (aka “theming”), proximity factors, distribution and co-occurrence, we feel that a certain measure of scepticism is in order here.

For one, Mr Garcia doesn’t seem interested in addressing any of the empirical data (read: experience) collected by thousands of SEO artists over a full decade, quite a bit of which runs counter to some of his assertions.

Second, while the theory of IR is a highly sophisticated field (and kudos to him for outlining this all-important aspect of it in such an easy to understand manner even for the non-academic reader!), this says nothing about what the search engines have actually implemented exactly in their ranking algorithms. Since they won’t divulge the technical nitty gritty of their operation for obvious reasons, most of his assumptions remain speculation, even if it’s a thoroughly educated one.

Third, the basic incompatibility between web design requirements, usability, human reading behavior etc. and fundamental IR parameters is not tackled or even mentioned. And that, of course, is where the #1 issue lies for any search ranking conscious webmaster, layouter and copy master.

Finally, on a more formal note, while the article features a set of footnote referrers, these are not included in the online version: a highly regrettable oversight we would strongly recommend the editors to fix at their earliest convenience.

So read → the full article here.

Subscribe!