Asking users what they want and aggregating those numbers provides data which can seem compelling and conclusive. However, it can also be a fickle guide to making good design decisions. At worst, the aggregate picture removes you from the specific motivations which influence users as individuals.

Kantar recently announced results from their survey of mobile phone retail buying habits. It is a worthy contribution to understanding the overall landscape of purchasing decisions and there were some notable conclusions:

• 30 percent do not research their purchase prior to buying.
• 63 percent receive a recommendation to buy Samsung while in-store, but 41 percent of those chose to buy something else. In contrast, only 3 percent receive a recommendation to buy Blackberry and 7 percent are recommended a Nokia.

(These figures refer to US customers, surveyed in Q1 2014)

However, there was one number which stood out, contained under the headline: “Smartphone buyers put functionality over design.” It stated just 7 percent of buyers consider ‘colour of the phone’ the most important driver of their purchase, while 43 percent cited ‘screen size’.

What are we to conclude from these numbers? It would be reasonable to focus future efforts on achieving the biggest practical screen size, while cutting efforts to differentiate through colour and materials. Kantar’s report goes so far as to state: “…it is clear that colour is not a deciding factor.”

Something about this doesn’t feel quite right though. The data, portrayed in a certain light, suggests a conclusive answer, but common sense encourages the use of additional tools essential to any designer: an ability to question the questions and explore existing answers from new perspectives.

Change the language around the data and a different picture emerges: more than 10.5 million US customers a year cite colour as the single most important factor when purchasing a new smartphone. The data is exactly the same – 7 percent of a US smartphone market estimated at around 150 million units in 2013 (source: IDC) but it encourages a different interpretation.

Perhaps it is good practice to always eschew percentages – which inherently remove us from thinking about individuals as individuals – and instead favour actual customer numbers. Charlie Dawson, partner at The Foundation (a consultancy which helps companies improve their customer focus), gave a memorable presentation at the very first MEX in 2005 in which he used the analogy of a dinner party to explain why percentages are a dangerous guide to improving customer experience. While a company might boast of a 70 percent satisfaction rate among its millions of customers, how would you feel if you organised a dinner party for 10 of your closest friends and 3 of them were unsatisfied at the end of it? You still have a 70 percent satisfaction rate, but you also have 3 unhappy friends.

Big numbers and percentages make pretty charts, but rarely get us closer to understanding user motivations.

Consideration must also be given to the competitive landscape. 43 percent said screen size was the most important driver of their purchase, but in a market where there are just a handful of screen suppliers and the size increases are now being measured in fractions of an inch, to what degree is being able to ship a 5.2 inch panel versus a 5.5 inch panel going to add value to your customer relationship?

Furthermore, the way in which these questions were asked will have different meaning for each user. When surveys are conducted at scale to generate statistically significant percentages it is by definition impossible to record the individual nuances which governed their answers. This should encourage us to question the questions:

• For instance, is there a chance users would not wish to seem frivolous by choosing something emotive like ‘colour’ over something practical like screen size when answering a survey, but actually behave differently when making a purchasing decision at retail?
• Or, if screen size really is the biggest motivator, does that mean bigger equals better on a linear scale, or is there an upper ceiling at which additional size becomes meaningless or even detracts from the product’s appeal?

Surveys such as this can be useful as searchlights to spot where opportunities may be emerging in the competitive landscape. However, each time the numbers seem conclusive, it usually pays to invest in additional research methods better suited to seeing the individual viewpoints of the actual customers which make up those seductively convenient percentages.