Log in

Login to your account

Username *
Password *
Remember Me
14 May 2015
posted by: Amanda

What shoppers think & what they do are two different things...

 think do

As the UK election pollsters found out, there's a big difference between what people think and what they do...

A non-client of ours in Australia sent me this link the other day with a quick note to say, 'this article made me think of you - showing the risk of asking people what they think/did versus observing what they actually do.  A good argument in support of the Shopperpedia methodology'. 

My obvious next question was to ask him again why he wasn't a client.... :-)... but that aside, it did inspire me to write this blog because what happened in the UK election, especially the pollsters predicting the result so badly, was... well... predictable.

At Shopperpedia, we live in the real world. At the coal face, where the rubber hits the road and that first moment of truth where real people (shoppers) make real decisions about whether to buy a product or not and there's one thing that we see consistently across all categories in the supermarket; there's a big difference between what shoppers think and what they do.

Starting with why this happens, we'll explain here why the UK pollsters got it so wrong and what you should learn from this in your role. 

Why in-store shopper research


So why does it happen?

There are five key reasons why what shoppers think they do and what they actually do are two different things. 

1. They can't remember
They bought 30 items in their last shopping trip, how can we really expect them to remember whether they noticed the promotional ticketing in soft drinks?

2. They never took that much notice at the time
FMCG is our life, not theirs and unfortunately (for us) buying groceries is not as exciting as buying a new pair of shoes or a new car.

3. They post-rationalise
Pasta sauce you say? Well let me think, erm no I have some in the cupboard so I definitely wouldn't have browsed pasta sauce in Tesco's yesterday... really?

4. They like to think otherwise
No, I definitely didn't browse Ice Cream because I'm watching what I eat right now.

5. They lie
No, I definitely didn't buy a ready meal as I'm a home cook and rarely cut corners.

wrong pano 13178

So why did the UK pollsters get it so wrong?

It was predicted by the pollsters that it would once again be a hung parliament yet in reality it wasn't that close (arguments about first past the post system aside).

To understand what people are going to do, it is not enough - if in fact it is required at all - to simply know how they answer questions in a survey. For all of the five reasons above outlined above, people don't tell the truth and therefore their opinion cannot be relied upon. The question, 'who do you intend to vote for?' can be likened to this case study and the question 'are you going to eat a bar of chocolate or have a glass of wine tonight?'. Would you know the answer when asked? If you did, (especially if it's a Monday night), would you answer truthfully? 

In order to understand how people are going to behave, you have to observe their behaviour. Let's look at an FMCG example to show how common data sources work together to give an accurate view - and prediction - of shopper's behaviour.

  • Scan data tells us if a shopper has bought Brand X today.
    Insight = they bought Brand X today. What's lacking here is insight into whether future sales for Brand X are at risk.
  • Panel or loyalty card data tells us that this same shopper also bought Brand X last time and the time before. 
    Insight =  they are a loyal Brand X shopper and therefore it is probable that they will buy Brand X next time. 
  • Online / out of store research tells us that when asked, they talk about how they love Brand X, it's the best and would never buy anything else.
    Insight = shoppers are very loyal and there is no perceived risk to future sales of Brand X
  • In-store shopper research observes that 20% of Brand X shoppers spend 50% longer at the shelf, that they pick up several other brands & actively compare packs at the fixture before putting Brand X in their basket or trolley.
    Insight = there is a perceived risk (20%) to the future of Brand X sales due to this loyal shopper's disloyal behaviour. 

In these commonly available data sources it is clear to see that an online or out of store survey about what shoppers think does not give the true picture. The same applies to the UK election pollsters. They got it wrong because the methodology focussed on what people think and not what they do. This case study shows the dangerous implications of basing core shopper strategies on such insights.  

data meaning

So what does this mean for you in your role?

As we've just read shoppers might well say that they don't go down the confectionery aisle when in reality it is one of the busiest aisles in the supermarket. Shoppers will also say that they are not influenced by promotions yet we see them comparing tickets and activation at the fixture on every shop trip. Shoppers will say that they are really interested in Australian made or made in England in an online survey or focus group but in reality they spend 10 seconds at the fixture and don't give the packaging a second glance. 

Or, more usually, they behave in an engaged way in several categories that they shop but just not in yours. Do you know why? What can you learn from these other categories and do better in order to change your shopper's behaviour and turn the conversion funnel around?

With retailers competing more and more for share and customer loyalty, every category, brand and manufacturer has their role to play in order to drive sales harder. This means having solid, reliable and appropriate insights at your fingertips. This means knowing what people do and taking what they think with a rather large pinch of salt.

Good luck!