By Marek Pawlowski, founder of the MEX conference.
It would be easy to assume the value of user studies is proportional to the sample size: identifying trends from 10,000 users will give you a clearer picture than talking to 10, or so prevailing wisdom would suggest.
However, what if the inverse were true? Is there an argument larger and larger sample sizes actually lead to business decisions which take you further and further away from delivering an experience tailored to each of your customers.
I was prompted to think about this at a recent meeting of The Foundation‘s Forum in London, entitled ‘Business needs data not intuition’ where leaders from a wide range of sectors, from banking to hospitality, were quite vocal in their support of using customer data for trend analysis.
Identifying trends within large data sets can be a useful aid to decision making when thinking about how to structure your business (e.g. how many blue widgets should you be stocking versus red widgets), but is far less effective at delivering something which actually meets the needs of individual customers (e.g. deciding whether a particular person is more likely to prefer blue or red widgets).
Customer preferences are infinitely complex, not to mention changeable. You might look at your user base and see clear segments of user behaviour, but no user actually identifies themselves as being representative of a segment: they are representative only of themselves and, everyone, when the data becomes granular enough, is unique.
Try asking a real person on the street which segment they belong to and I’ll be surprised if they don’t ask you in return whether you’ve gone mad.
The mobile industry provides customer experience practitioners with an interesting testing ground for developing products tailored to ‘segments of one’. Each device can be uniquely identified and an increasingly detailed set of contextual data can be gathered about each user.
Married to the latest database technology, the two-way communications links and growing number of contextual sensors in mobile devices allow every nuance of customer behaviour to be understand.
The problem, however, is that most business models are incompatible with responding to users as individuals. Companies are still built around ageing principles such as economies of scale, mass production and distribution. Within these frameworks, it is far more convenient and cost effective to group behaviours into segments, thereby moving you further from the reality of what each customer wants as an individual.
Of course, there are limits to what can be achieved with current technology. Today it is simply uneconomic to produce phones personalised for each customer at the point of manufacture – but this is changing. The percentage of the mobile experience abstracted into virtual, customisable elements is growing daily: what staretd with ringtones and wallpapers has expanded into applications, user interfaces and input methods. With haptics, even the tactile qualities of the device be personalised. In the future, the colour of the exterior casing will be virtualised using new skin display technologies.
The challenge for businesses is to build their architecture with a long-term goal of being able to make each customer feel as if they are receiving a service tailored to them as an individual. This infinite personalisation will become possible by degrees over the next several years and companies who put the foundations in place today will find their advantage grows daily as they become better and better at meeting customer expectations and competitors who did not plan for this shift find it harder to respond.
It would seem, therefore, that more data are the solution. However, businesses must ensure they can ask different questions to every customer and respond to their answers as individuals.
Question: Which methods does your business rely on to understand customers as individuals? How do you address the challenge of translating that understanding into an individual dialogue with each user? Will the mobile industry ever truly be able to offer a personalised service for customers? Post your comments on the blog…