Web site personalisation - or information intrusion?

I’m trying hard to get a handle on whether the new breed of personalisation and recommendation technologies currently spawning across the web are likely to be in the, “oh my god how did we ever live without that?” category - or simply another intrusive and invasive potential fissure for security concerns.

Amazon already knows which books we buy of course and likes to offer us more of the same, Google mail sometimes has some infuriating link options based upon what my previous searches might have been showing somewhere near my inbox – but thankfully I too busy reading mail to bother noticing these most of the time. But what we’ll see next, or so I read, is technology that treats individuals on the web as more ‘unique’ (ouch – sorry, that slipped in) so that recommendations are delivered at the individual level and are based on a combination of historical information and real-time shopping behaviour.

Bubbling up on the developer newsometer are the likes of companies that “do” this kind of thing such as CleverSet. This is essentially advanced statistical relational learning based on Bayesian network analysis and was originally developed by the company for the US Department of Defence.

This approach is said to drive more relevant experiences for online customers by using more types of data from more sources (such as time & date and seasonality), compared to techniques like collaborative filtering.

CleverSet would of course have us accept the benefits of product recommendations, direct email campaigns, recommendation-driven search results pages, personalised presentation and customised content – they say that recommendations appear on the web pages and in the locations that you, the user, have specified.

Personally, I love my iGoogle personalised gadgets home-page and welcome the chance to read news headlines, quote of the day and my inbox all in one go. So why not have more of my web experience personalised? Can we be bothered? Do we have time? Aren’t we all used to interpreting search results the way we want to by now? Will this be personalised perfection or will it be information intrusion?

In a December 2007 report entitled “Which Personalisation Tools Work For eCommerce — And Why,” Forrester Research principle analyst Sucharita Mulpuru noted: “Web site personalisation is an underleveraged weapon… As the amount of content online grows — particularly on e-commerce sites that often sell tens of thousands of different products — and consumers are left to wade through cumbersome experiences on their own, personalised e-commerce experiences promise customer engagement and loyalty through increased relevance. Enabled by external tools sometimes called personalisation engines, recommendation engines, discovery engines, or behavioural targeting tools, personalisation allows retailers to increase relevance through activities like matching cross-sells to customers based on interests or customising clickstream paths based on previous purchase or visit histories.”

Whatever your opinion, Art Technology Group last week acquired CleverSet for $10 million in cold hard cash, so they clearly believe in these concepts… so we may very well see this strain of technology surfacing in the stacks’ of the web elite very soon. Or will we?