Software is slippery stuff

Like most knowledge based industries the IT industry sometimes throws up wonderful scams that exploit the ignorant and extract cash from the unwary, remember the millennium bug? One such honey trap currently popular with marketing departments around the world is that of "SEO" or Search Engine Optimisation. The problem people want to solve is how do I make sure that my WEB site is listed on the first page (or as high as possible) in Google when people search for things that are relevant to what I sell?

There are two ways.

1. Pay Google
2. Fiddle around with the contents of your WEB page so that the Google page-rank algorithm (a piece of software that assigns a score to your page relative to other pages for certain keyword searches) automatically ranks your page highly for a desired set of relevant keywords.

The only problem is that the Google page-rank algorithm is secret and mysterious, the only certainty is that it's completely opaque to non-IT literate consumers. Consequently there's this whole industry that's built up around the idea that there are certain "things" that you need to do to structure your WEB pages so that Google thinks they are worthy of a high page-rank, this is of course true at a simplistic level, but the supposed "things" you need to do have long since entered the realm of mythology; the other problem is that Google change their algorithm from time to time supposedly to keep everyone on their toes.

The task of building a WEB site is complex, the task of building a WEB application even more so, however the number of variables that the GPRA (Google page-rank algorithm) has to work with is relatively small and hasn't changed much in 10 years. As is often the case common sense seems to provide the most reasonable guide to what is needed, the crucial thing to understand what Google is trying to achieve through their page-rank mechanism, this strategy is simple. Google needs to deliver pages that are most likely to be of interest to the provider of a search term. Why? because if consumers don't get useful content they don't use Google and if that happens then they don't read their "paid for" advertisements alongside that content which provides revenues to Google, not rocket science, its a "give to get" model.

So how do Google work out which pages will be most useful, clearly a number of parameters apply, first how many other people also find the content useful? If lots of people think a piece of content is good then chances are... it's good. So a prime feature of the ranking calculation is to look at how many in-bound links there are, i.e. if someone else links their page to your page then that's a good thing, if their page is also popular then even better. This was a fundamental part of the early page-rank scheme however just like in nature the "organism" that is Google page-rank is engaged in an evolutionary arms-race with the predators and parasites out there that wish to exploit it. Once people figured out that if you have a lot of inbound links to your content then you get a higher page rank then they simply set up link farms that proliferated bogus inbound links. Unsurprisingly Google now penalises sites that are obviously linked to via link farms. Similarly if a page mentions the search term that the consumer is looking for a lot of times then chances are that the page must absolutely definitely be about that subject, right? Wrong!, as soon as people discovered that loading pages up with keywords improved their ranking Google changed the rules so that only the "right" number of links in the "right" places (i.e. visible to the consumer) scored highly.

In the end there is only really one guaranteed way to ensure that your pages are highly ranked and its the most obvious thing to say and yet one of the hardest to do, i.e. create good content that lots of people want to read! Any other "quick-fix" or "silver bullet" (no matter how much the SEO vendor is plugging their latest "voodoo") is at best temporary and at worst a waste of time and money.