Never mind the quality, feel the traffic

7 September 2005

Given a choice between two metrics, most people will take the one that is easy to measure and one that is more accurate, but difficult to capture, which one do you think most people would pick? That's right: ease of use wins. Which is why I am less surprised than the chief of search marketing firm iProspect, Rob Murray, about the results of a survey his company asked JupiterResearch to perform.

When asked how they were evaluated by their employers, 80 per cent of search marketers said their performance was tied to some search-engine marketing metric. Half of them looked for increased traffic, 40 per cent for more clicks through to their sites and just under half cited search-engine ranking. The more difficult to measure criteria, such as awareness among consumers or even actual sales - things that any advertising-related campaign should be promoting - came some way further down the pecking order.

The survey was among companies with dedicated search marketing staff apparently, which suggests that the companies involved are quite advanced in trying to get their sites up the Google and Yahoo rankings. That does not seem to make them any more advanced in working out just why they might be doing it, however. It would be interesting to know the results for a survey of clients' attitudes to external consultants: I can't imagine it will be much better. And this is not good for the health of the Internet as a whole. Narrow, non-financial metrics that are not tied directly to what the business needs are just so easy to fake. At least the search-engine people have some ways of detecting click fraud: is that level of traffic analysis happening among companies with websites?

If I had inclinations to dabble in the black-hat SEO business, I'd be ordering me some bot farms and hammering clients' sites hard. "You just want traffic? That shouldn't be a much would you like? Mild interest, a BoingBoing/Slashdot server burner or denial-of-service levels?" Similarly, the high page-rank power of blog links will only encourage the black-hat gang to step up their spam blog-making activity. And who can blame them if the metrics that companies use are so poorly structured?

Even the iProspect press release that popped up on PRNewswire has been drafted into splog-land. The top results in Technorati - in fact the only results - for it when I last looked were all from splogs designed to pump up page ranks for other sites or just generate clickthroughs via AdSense.

The sad thing about this focus on bad metrics is that it is a feature of the wider advertising space. Because the people booking the ads do not bother with difficult metrics, poor-quality media often end up squeezing the good out of the market because they focus relentlessly on the factors that advertisers use as metrics. Media that people actually see and read often suffer badly when compared on measurements that focus on narrows aspects of campaigns rather than wider results. The new techniques such as search marketing seem to be following the same old patterns.