
I think one of the most potentially misleading questions that we can ask in a survey is “do you plan to increase/decrease your spending on [blank]” for household items and packaged goods. I honestly think that the responses given to these kinds of questions have a very tenuous relationship with actual behaviour.
One of the reasons for this is because these questions do not take into account the mental accounting that people use to budget. People do not think of their spending by individual items – unless it is cars or holidays or large ticket items. They group things together. Within these groups they subsititute. And the groups themselves are dynamic. For example – people are unlikely to plan their future expenditure on ice cream. They are likely to consider “groceries” and there are a myriad of ways in which they could calculate their spending on that – by month, by trip etc. Research we conducted on alcohol found that young adults simply exclude beer from mental budgeting! Of course when prompted they may give a response that they plan to cut back but the fact is they hadn’t really considered it until we asked the question.
So when a research asks “do you plan to increase/decrease your spending” on an individual item this may be the first and only time they really think about it. The fact is they have no “plans” regarding this item. This is especially true if we are asking about their plans for the future. I do not believe that a consumer can accurately tell us what he or she plans to do about spending on alcohol or ice cream or paper towels “after the recession”. And even if they do I do not think their response reflects what might happen – which can depend on their ‘mental accounting’ and where that category fits into their budget as well as a myriad of other changes – channel preference, occasions etc.
The right way to do this is to understand their mental accounting process, the heuristics they use for budgeting and which bucket the category falls into. And of course you need to also understand circumstances such as changes in channel choice and occasions. He problem is of course this requires a lot more questioning that “do you plan to …”. But hey, if you want to get it right