The value of personalisation

by Paul Boag

I have always been vaguely suspicious of the value of personalising content on a web site however a recent survey has thrown up some interesting figures.

Disclaimer
I feel like I should point out at the outset that this survey was conducted on behalf of ChoiceStream, who are themselves a supplier of personalisation capabilities for web sites. This makes you question the reliability of the information it contains. Nevertheless I found it fascinating reading

Background
The survey was completed by 673 respondents in May 2004. In order to track changes in consumer attitudes over time, the survey is continually repeated at regular intervals.

A major contradiction
Overall, the survey found that more than 80% of consumers were interested in receiving personalised content. I found this figure much higher than I expected. What did not surprise me is that the percentage of people willing to provide information on their preferences and interests in exchange for this personalisation was much lower at 64%. Users continue to be reluctant to hand over personal information of any description however good the reason.

The problem still remains of how to present this functionality in a way that doesn't require too much user interaction or give the perception of being too invasive of privacy.
With a significant proportion of users unwilling to provide information to help personalisation the options are extremely limited. The only feasible way forward would be to monitor users behaviour and provide personalisation based on this. However when asked if they would be willing for this type of monitoring to take place only 40% of the sample group agreed. Seemingly people are even less happy to be monitored than they are to answer questions.

Conclusions
All in all this makes frustrating reading for those running web sites. On one hand users are saying they want personalisation especially in the area of ecommerce and yet on the other they are unwilling to provide information or allow monitoring to make this happen. However what a user says they are willing to allow is very different to what they except on a day to day basis. I suspect that if the question had been "are you happy for Amazon to recommend books?" the response would have been very different. The fact that Amazon monitor user behaviour to achieve this would not worry or even occur to users.

This survey has made me think again about personalisation. I am still far from convinced of its value in many situations. Just because technology allows us to do something doesn't always been it is worth doing. However in certain circumstances it has real benefit. The problem still remains of how to present this functionality in a way that doesn't require too much user interaction or give the perception of being too invasive of privacy.