Consumer Behaviour

One of the advantages of understanding how the process of asking people questions influences them is that you can have fun demonstrating its impact.

I was recently invited to speak to a local business group and took the opportunity to demonstrate the frailty of asking questions and the nature of the unconscious mind.

Whilst the samples were too small to be scientifically valid, the differences in responses to my fake research were both predictable and entertaining.

I set up a taste test using three very similar products: one was from a value range, one from the ‘standard’ range and one the premium offer.  Everyone was led to believe that they were taking part in legitimate market research and that they all had the same questions.

In fact there were five different questionnaires, all asking the participants to taste three products and answer some questions.

What was I able to demonstrate?

1. People will express a preference when given identical products to rate.
2. People’s taste preferences are influenced by branding.
3. People have no clue how much most products normally cost.
4. When asked to analyse aspects of a product (like sweetness and texture) people reach a different conclusion about which product is best (compared to when asked simply to select a favourite).

Philip Graves consumer behaviour, market research market research

The title the ‘International Journal of Market Research’ (IJMR) sounds undeniably impressive.  Generally speaking journals are good things, bringing together peer reviewed papers from people pushing the boundaries in a particular field.

But I wonder… do astrologers have a Journal of Astrology?  Google suggests that they do, sort of – it looks as though it might just be one astrologer selling predictions.

There’s a National Journal of Homeopathy – I wonder, to paraphrase Tim Minchin, if they’ve had any papers on how water forgets about the wee and poo it’s had in it and just remembers the traces of medicinally advantageous ingredients?

My point is that it’s easy to get a false sense of validity from a name.  In Consumer.ology I describe market research as a pseudo-science and, arguably, having an ‘International Journal’ is all part of the industry’s mystique.

I must declare a personal interest at this point: recently the IJMR reviewed my book; the reviewer hated it.  But the review, a viewpoint article and one of the papers made for particularly interesting reading: let me explain…

[Curiously, the reviewer's reading skills didn't extend to my name: I was called Robert Graves and Peter Graves, but never Philip.  I wonder if the book had him in such a rage that he struggled to focus on all the words.]

Pedantry aside, the IJMR’s choice of reviewer was somewhat self-satisfying.  As far as I can tell, the reviewer earns his living conducting the research that Consumer.ology questions the fundamental validity of.  How is such a person to address the evidence from psychology and neuroscience that shows people aren’t very good witnesses to their own behaviour and that the process of asking them what they think influences them to say particular things?

The somewhat strange thrust of the review is that the market research industry knows all about the problems that lead me to believe asking questions isn’t worthwhile – “… they have long been widely recognised by many in market research.”  Strange then that if you look at the Market Research Buyer’s Guide virtually every single one of the companies listed is offering the sorts of research that is beset by these problems.

Do they tell prospective clients all about these “widely recognised” problems when they come to them requesting a survey or focus group be commissioned?  I suspect not.  Partly I suspect this because no research company has ever mentioned them to me when I was commissioning research.  And partly I suspect not because, were they to mention the problems, the research wouldn’t go ahead.

Entertainingly, the review was, I assume, written before the recent Scottish Parliamentary elections.  My rejection of opinion polls was described as ‘…simply not justified by reality.’  He adds that, ‘When it became clear a few years ago that something was going awry in the accuracy of such polling, the industry effort aimed at addressing these issues was remarkable…’  Remarkable until they got the results wrong again in the Scottish elections.

Election polls conducted close to the date of an election should be quite accurate, and yet even they prove problematic to the research industry.

Elsewhere in the same edition two contributors attempt to reconcile market research with behavioural economics.  In a ‘Viewpoint’ article Nick Southgate attempts to align asking questions with behavioural economics by pushing it into the gaps that a behavioural approach can’t fill. That behavioural economics can’t identify (at least not directly) the content of people’s decisions, doesn’t mean that market research can.

In a (presumably) peer-reviewed paper (market research peers, of course), Wendy Gordon draws the astonishing conclusion that a new branch of (dynamic) market research can help provide behavioural economics with “the practical skills and applications that they need to solve the problems that face them in an increasingly complex world context.”

Psychology and behavioural economics have provided the basis for identifying the folly of traditional market research (evidence that is quoted in the paper): expecting the people whose work up to this point has been so clearly undermined to be the custodians of a new approach is, it strikes me, somewhat risky.

This risk is evidenced when Gordon advocates “not throwing the baby out with the bath water” before going on to suggest that qualitative research has strengths in certain areas including identifying values and beliefs: behavioural economics has demonstrated that espoused values and beliefs can be (and frequently are) irrelevant, when examined alongside actual behaviour.  Surely, that’s a case of pouring the bathwater back in again.

The problem for the IJMR is that people don’t approach anything with an open mind.  We arrive burdened with associations and experiences that colour how we interpret what we find.  It seems that the IJMR wants to take the challenges from outside its field and force its existing techniques onto that evidence.  Put another way, how can someone who has spent their entire career asking questions reconcile themselves to the information that there is little evidence to support the validity of asking questions and much to undermine it as a reliable tool?

The belief that you can ask people questions and what you hear back will be an accurate insight is just that, a belief.  It’s an apparently plausible concept that fits with people’s view of themselves as the conscious agents of their actions.  Its reinforced by those times research appears to confirm something we believed, or appears to be born out by what happens next: of course, the same benchmarks are what perpetuates the use of things like astrology.

With so much that is central to market researchers’ beliefs questioned in Consumer.ology, I would advise them not to read it.  Or at the very least, they should consult their horoscope to see if it’s a good day for reading a book.

Philip Graves

Philip Graves consumer research, Getting a Book Published, market research consumer.ology, IJMR, market research

Every now and then I receive an invitation to complete an on-line survey.  They’re normally hopelessly poor tools at understanding real consumer motivations.  They interrogate the wrong part of the respondent’s mind (the conscious mind) and unwittingly influence the part they should be targeting (the unconscious mind).

So today, I thought that, as I plod through the survey, I would include a running commentary of what’s bad, just in case anyone else is thinking of running an on-line survey any time soon. 

The First Few Questions
A few classification questions to see where I live, when I was born, to check that I watch television (the subject of the research) and to check that I’m not involved in marketing, market research, journalism, advertising, public relations or television: arguably I’m involved in all of these but, just like any other respondent, I’ll be ticking the answers that suit my purpose: “None of these”.

Questions
The first proper question asks me what channels I can name, other than the five original terrestrial channels.  This awareness question has evidently been written without any awareness that memory is context-based.  Sitting in front of my computer I may recall one set of channels, but when it comes to watching a programme I will select in a completely different environment and mindset.  Has anyone correlated top of mind channel awareness with anything?  What, you many well ask, will this question prove?  I suspect that a TV channel has been advertising itself and thinks that unprompted awareness is a good measure of their advertising impact; I’m betting they can’t substantiate a link between the two.

I’m now presented with ten channels and asked to say how often I watch it, of if I never watch it, if I’ve heard of it.  My options are “most days”, “at least once a week”, and so on.    I select a programme by the programme’s title, I don’t always know which channel I’m watching.  Fortunately, the television companies have tracking data from boxes that actually monitor what a sample of UK viewers watch: why, you may well ask, are they inviting in this meaningless self-reported data?  Presumably it will drive questions later in the survey, but since I suspect I don’t know which channels I watch when I flick through, this isn’t going to be very accurate.

I’ve just seen I have to do this for sixty programmes; who would volunteer for this kind of pain normally?

Now I’m invited to say “how I feel about each of these channels” ranging from a “channel I love” to “no strong feelings”.  So I will rationalise my feelings about a channel.  What, you may ask, would it say if I watched a channel a lot that I didn’t love?  Or if I love a channel but don’t watch it often?  There is an implicit assumption in the questionnaire – in my case one that is misfounded – that I know what I’m watching and can match my viewing to a channel. 

Oh Dear
I realise this might seem like a writer’s licence, but I have actually been going through and documenting my thoughts step by step.  Having reached this question about “channels that I love”, the on-line survey has broken.  I’ve tried two browsers and several refreshes, but the page doesn’t work.  The table of pre-coded responses and channels has disappeared.  Nothing I click helps.  Going forwards delivers a warning that I need to give my answers, but I can’t because it won’t let me.

So, regrettebly, I can’t get to the heart of the survey.

But already I know that no one should attach much significance to this survey’s results.  At it’s heart is a fundamental assumption that people link their viewing to the channel they are watching.  A lot of TV viewing is done in a very ‘withdrawn’ mental state.  We let a familiar programme wash over us like a sort of brain support system: there’s processing going on, but it requires no effort or involvement from us.

Ask yourself this: what did you watch on television the night before last?  The chances are that, unless it was a significant sporting event or a film that you had actively selected, you won’t remember the titles of the programmes, never mind the channels.

I’ve been commissioned to study viewing behaviour and programme selection behaviour in the past, and neither has very much to do with the conscious mind. 

As always with poor market research, the waste in terms of the money spent on the survey is probably minor in comparison with the cost associated with making the wrong decision and not acquiring

Philip Graves

Philip Graves consumer behaviour, market research commissioning, market research, on-line survey

Developing a logo is an interesting experience.  Recently several organisations have found that the internet provides a platform for dissenting voices to grow into active movements to oppose designs that they don’t like.

Gap, who some have suggested were really undertaking a publicity stunt, Starbucks and the Portland Timbers have all experienced an adverse reaction when plans of their changes came to light.

Recently, the Iranian Olympic Committee has said that they think the London 2012 logo is racist, spelling out the word Zion.  There are a number of reasons not to pay too much attention to this complaint:

1. It doesn’t spell out Zion, it says 2012.
2. If it did spell out anything in English it would be “Zo in”, since we read left to right and, with no hyphen, the second line should be taken as a new word.  Perhaps animal welfare groups should be boycotting the Olympics instead!
3. If you want to say that a 2 looks like Z, is it reasonable to also claim it looks like an N?  So that would mean it spells out Zoin.

Choosing a new logo is hard.  I should know, I’ve just been through the process for my website (it should be appearing there in the near future).  It’s tempting to outsource responsibility for deciding what design to choose to other people; I could have asked my friends or conducted a poll with people who have signed up at my website.

But if I don’t know what conveys my brand, what perspective are other people expected to respond from?  They might tell me what they think looks nice, aesthetic taste is understood to be intrinsically personal.  They might tell what they think my brand is about and what conveys this to them. 

But this is my chance to communicate a little bit more about me.  If I already convey everything I want to, if the status quo is so important, why would I be creating a new logo at all?

One comment famously suggested the London 2012 Olympic logo looked like Lisa Simpson performing a sex act; lots of people agreed.  But they only agreed when one person said it and that comment was picked up by the media (at which point its humour and astute observation caused it to spread as only a meme can).  I don’t doubt that, had the designers had this pointed out to them, this design would never have seen the light of day.

However, just because, once it’s pointed out, you can see it, doesn’t mean it would ever occur to you to think along those lines independently.  Consequently, you could ask one million people for their views on the logo, but if you don’t ask the person who makes that specific association your research will be irrelevant (assuming the media find and propagate the comment subsequently).

It’s one thing to run a logo design past a fresh set of eyes, to check that you haven’t inadvertently conveyed something directly that is offensive to people who will see it, but it’s reckless to let a consensus market research opinion drive the representation of your brand.

I often think researching a logo is a little akin to asking someone to dress and behave a certain way when they go on a date with you: in theory it should make for a perfect evening, but the fact that someone is just playing back what you want isn’t a substitute for the genuine experience of two people being themselves and enjoying each other’s company.

So by all means run a design past a few pairs of eyes, and ideally have people from different cultures check you aren’t offending them inadvertently, but don’t suppose that other people can tell you how you should be.

For the record, I have always liked the 2012 logo, I never liked the Gap logo, I like the new Starbucks logo and the new Portland Timbers logo appears to me a substantial improvement.  But my view on these doesn’t matter.  And whilst I hope you like my new logo when it appears at my website, if you don’t, well, perhaps it was just never meant to be between us!

Philip Graves

Philip Graves market research 2012, design, logo, market research, Olympics

People aren’t desperately honest creatures.  Through no fault of our own we’re victims of the way our brains have evolved; it’s wise not to take the things people claim at face value.

Among the many issues affecting market research the quality of respondent recruitment is reasonably frequently debated.  It’s not something I got into in Consumer.ology mostly because even when you recruit the “right” people, asking them questions throws up a whole world of other issues.  

However, over the last couple of days I’ve had a fascinating insight into the recruitment process and can, at no charge to the market research industry, offer them a high quality recruitment tool. 

I was contacted by a television network who wanted to interview me about a story that has been in the news regularly over the past few weeks; the cost of filling your car with petrol (or diesel).  Prices have risen substantially over the past few weeks and since January 2009 the cost of filling your car has increased by almost 50%.

The television network wanted to talk to me about consumer behaviour, they also wanted to interview someone who had changed their driving behaviour as a result of the soaring prices.

Having heard lots of people complaining and discussing the issue, not to mention talking about what they were going to do differently - let’s face it, it hurts to refuel your car these days – I thought it would be easy to find someone who would talk to a journalist about the changes they were making.

So I sent out an email to a hundred or so people locally; people encompassing a wide socio-economic spectrum.  The researcher working on the programme contacted various other people and organisations to find someone.

The response?  Not one person was willing to stand in front of a camera and explain how their behaviour had changed.

Now, of course, not everyone likes the idea of being on television, but given the concern in the media about this issue it would be easy to presume that people are doing something differently.  Then again, the story “Petrol prices are skyrocketing, but people are still buying it like usual” wouldn’t be much of an attention-grabber, would it?

Talk, as they say, is cheap.  Airing your anxieties, thoughts and feelings is pretty much an everyday occurence.  But actually accounting for your behaviour, and feeling that you might be held to account for your behaviour, is quite a different matter.  [This is one of the reasons that when I do ask questions I use interview techniques that don't allow people to talk about their thoughts and feelings.]

With the jury of your peers in the pub, who see you on TV talking about the fact that you now “cycle a lot more” or “take the bus instead of the car” or “can’t afford to drive”, there is someone on hand to poke you in the side and laugh at your fanciful exaggerations (or at least you are afraid that there might be).

I suspect that the conscious introspection triggered by realising that they would be accountable for their answers was at least a contributory factor in people’s unwillingness to speak to the press.

So, when it comes to market research recruitment you should probably recruit people on the basis that their responses will be broadcast to the nation, that way you can be more confident there’s some substance behind their claims.

Talk is cheap, research talk is even cheaper, but claiming you’ve done things you haven’t in front of your mates can be expensive.

Philip Graves market research consumer behaviour, market research, petrol prices, research recruitment

With my book Consumer.ology now published I’m starting to hear back from people who have heard about it, read it or read or heard an article or interview about it.

One of the very positive upsides to this is that more people are starting to share their stories of market research getting it wrong. 

Whilst I managed to unearth a good number of examples for the book, the fact is that it’s not really in anyone’s interests to publicise occasions when money spent on research was wasteful.  Occasionally there’ll be times when someone’s decision was vindicated and they’ll speak about it, but often the people making the final decision are also the ones who have decided to spend several thousand pounds on research, and choosing to ignore it doesn’t reflect particularly well on that decision even if it’s the right thing to do!

One reader contacted me to tell me about the conclusions of a focus group for a beer that was being tested with consumers in the UK, with a view to importing it.

The conclusions from the research were that the beer was “weak”, “watery”, “gassy”, “… like kissing your sister!”.  It certainly wasn’t a real man’s beer.

However, the autocratic head of the business decided to push ahead with launching the beer in the UK.  His name was August A. Busch III; you may have seen the product around here in the UK since then… it’s called Budweiser.

If you have any market research stories you would like to share, please do get in touch (philip(at) philipgraves.net)

Philip Graves

Philip Graves market research Budweiser, consumer.ology, market research

Yesterday I visited the Marketing Week ‘Insight Show’ at London’s Olympia Exhibition Centre.  I was underwhelmed.

For the most part the market research companies who were exhibiting were pedalling the same old methodologies in the same old ways – all the stuff that doesn’t really work but makes companies feel better.

A number of firms were presenting themselves as excellent at helping clients with innovation: I’m not convinced that this is something market research can help with: as Henry Ford said, “If I’d asked people what they wanted they would have said a faster horse.”  Or perhaps they would have said they wanted the same horse cheaper.

I have no problem with talking to consumers – it’s a very worthwhile thing to do: there’s a small chance that one person might say something that sets an idea off in your head and leads to create something amazing.

But the problem with market research is that it thinks it can find consensus: it can’t resist asking people to come up with ideas and then gauging how interesting people think those ideas are.  That’s pretty pointless.

Firstly, a concept expressed as words can take on a myriad of forms once developed and it may be the form that determines whether consumers take to the idea or don’t.  No one would have said they wanted an iPod: there were already MP3 players around that performed the same function or that were technically superior: but none of them looked half as good as the iPod did.

Secondly, people are hopeless as predicting what they want in the future: context is perhaps the single biggest influencer of behaviour, and if you see a product everywhere you’re far more likely to want one.  If a friend tells you it’s great, that can make you want it too.

Thirdly, for the average product to be successful how many people out of every hundred need to like it enough to buy it?  If you have a group of eight people coming up with ideas it probably only needs one of them to love it for it to be viable – but who launch a product that seven out of eight people hated?

So, by all means talk to consumers, get them coming up with ideas and listen to the things they say spontaneously when they are buying your products or talking about them naturally if you can: but don’t pay thousands of pounds to a market research agency for this. 

In fact, don’t let somebody else do it.

Do it yourself.  Do it with your product experts and product designers.  These are the people who have the base knowledge and existing expertise that might be sparked into a new dimension by what one person says.

And, for the most part, as Henry Ford realised, it will be ideas from the experts and enthusiasts inside a company that will see the opportunities and develop the best innovations.

For example, perhaps you ride a bike.  If you do, I sincerely hope you have a bike helmet.  If you don’t, please get one!

I expect you think your helmet is fine, that it fits reasonably well and will do the necessary work in an accident: I hope you’re right.

But, the chances are, that your’s doesn’t fit very well.  You just don’t have a good fitting helmet to compare it to, because most helmets don’t fit well.

Yes, they’re tested; but they’re tested on dummies for impact protection not people with funny shaped heads who sweat and move around.

Yesterday I discovered the Kask brand of cycle helmets.  They fitted my head like no other.  They stay in place.  They don’t slip up if you gently push the front.

I looked around on-line and found out that lots of other people had discovered the same thing as me.  I also discovered what I consider to be the best product review I’ve ever read.  It was in the comments at www.cyclosport.org after the website’s official review.  Some chap called Mark Liptrot said the following:

“£155 for a piece of polystyrene that the excellent innovators at Kask have had the genius to make head shaped.”

Sometimes innovation is simply about not settling for the status quo: consumers aren’t going to help much there because they, for the most part, rather like things as they are.  But when you show them things can be better, you can charge rather a lot for not very much and they’ll be happy to pay.  My Kask helment will be arriving soon.

Philip Graves

Philip Graves Innovation, market research innovation, kask bike helmet, market research

It’s often claimed in market research circles that perception is everything.

I suspect this stems from the thought that, provided the customer perceives things as being good then that is what matters, be it customer service, product quality, your brand’s image.

I’ve come to the conclusion that this is a really stupid way of thinking about consumers. Let me explain…

In one sense, perceptions are all that customers have to provide in market research. Life isn’t absolute when you’re living it. For example, I’ve been having some lower back ache (too much time spent writing in a bad chair) and have been treating it with a physiotherapist and lots of exercises. Unlocking my lower back has caused other parts of my back to react and the other night I turned over and stretched at night and managed to pull a muscle higher up my back.

Now, at the moment, my lower back doesn’t hurt. So I perceive that my lower back is fine. However, this is largely because my pain attention is directed higher up at the part that really hurts. In reality, the issues in my lower back are not resolved yet, it’s going to take a few weeks to undo what I’ve done over many months. So my perception about my back is almost certainly misleading.

If you ask me now how bad my back health is I will say it’s fairly bad: my attention is directed at the pain right now and I’m very sensitised to backs.

But if you ask me in a couple of weeks, when I expect it to be considerably better, I will probably say that my back is pretty healthy. In a couple of weeks it shouldn’t be hurting me any more, but I’ll have many more weeks of exercises to do to make it properly mobile.

Perceptions change over time. The further people are from an event the less likely they are to perceive it accurately.

When it comes to customer service (for example) what matters is what service you actually deliver. The only way you can judge that is by seeing what experience you deliver to customers.

They may say that service is poor because they’re expecting red carpets and rose petals.

They may say that service is good because they have such low expectations of your brand.

You have no way of knowing what frame of reference they’re making, how time is shaping it or how recollections of experiences will be factored in to future purchase behaviour.

All you have is a score. Which feels compelling because it has a number attached. But it’s actually meaningless.

Ah, they say, we can track it over time.

So now you know how a meaningless number changes. That’s helpful!!

The solution?

Decide what quality of service you want the customer to experience and design your service to deliver that.

Then watch, dispassionately, to see if it looks as though customers are getting that experience.

If you want them to leave the store delighted, look for delight on their face: we all know what ‘delighted’ looks like, don’t we?

If you’re studying the moment that customer service happens you’re studying reality. If you’re asking customers what they think you’re getting perceptions but you have no way of knowing their durability or relevance to that customers future behaviour.

There. That should save your company some wasted money on customer satisfaction surveys.

Philip Graves

Philip Graves Customer Service, market research Customer Service, design, market research, perception, reality

So here’s the thing.

If I convinced you that crossing the road without looking was a wonderful, amazing, liberating, life-transforming experience, and I mean really convinced you so that you believed it, what would happen?

You would take your first attempt with extreme nervousness, shaking with fear, but you would probably make it across the road just fine.

With your senses heightened you wouldn’t really be using your new-found faith in traversing streets with total disregard for what was passing: no you would heighten your other senses until you were certain it was safe and then over you would go.

Of course, after a while, if I’d done a really good job and kept reinforcing the message, you would start to foster the genuine belief that you didn’t need to look to cross the road.  After all, you’ve now got my enthusiastic message – and I’d tell you about how many millions of followers were doing it; you’re own ‘experience’ and, in all likelihood, you would have started to notice other people who didn’t seem to be looking when they cross the road (confirmation bias).

For a while you might have luck on your side too.  Most car drivers aren’t looking to run a pedestrian down.  They would drive around you or stop their vehicles.

Their anger would surprise you.  But you would smile inside, knowing that they simply didn’t have your deeper understanding.

There would, however, almost certainly come a point at which you would get hit by a vehicle.  It would hurt, a lot, and you would be extremely confused.  But would you abandon your belief that you don’t need to look when crossing the road?

I suspect that you are telling yourself, “Of course I would!” and probably also, “I would never have believed it in the first place.”

The latter is almost certainly true.  But why?

You haven’t (I hope) experimented with both techniques.

You believe you should look when you cross the road because it’s been drummed into you as a child.  Quite possibly with a good explanation that seems eminently sensible.

But you don’t KNOW for certain because you haven’t personally tested all the alternatives (and, incidentally, please don’t start exploring this theory with road crossing!).

If you really believed in not looking when you crossed the road, you would almost certainly carry on believing it.

Don’t believe me?

Let’s take a less extreme example: Market Research.

In the last week I have listened to or read two discussions on two unrelated surveys.

Survey one explored a number of questions relating to people’s bank accounts.  It asked people if they had moved their main account recently and also if they were happy with their bank.

Now, to set the scene: Here in the UK the banks are coming under huge pressure from some consumer groups and government regulators.  It’s been noticed that people aren’t switching to get the best deals and the finger of blame is being pointed at the banks for not providing customers with information and for making the process of switching from one bank to another extremely difficult.  Recently there was a mass action court case claiming that the charges banks levied were unreasonable.

The survey discovered that 93% of the people who hadn’t changed banks in the past two years (which was 92% of the people interviewed) were “happy” with their banks.

One reporter was “astonished”.  They even had a psychologist (sadly not me) to explain why the result might have been so high.  She pointed out, very reasonably, that having asked if they had changed banks and stated that they hadn’t, it would have been difficult for them to say they were unhappy.  She also pointed out that “happy” is a word that is open to a multitude of interpretations: aren’t they all!

So does this mean they’re happy?  Or does this mean the questions have primed them to say they are happy when they otherwise wouldn’t have done?

The answer is both.

The second survey, an internet poll, concerned the strike action of British Airways cabin crew.  It asked: “As the world’s highest paid cabin staff, should BA cabin crew go on strike?”

The result was that 97% said “no”.

Is the question fair?  It seems extremely leading, doesn’t it.  But if the statement about them being the highest paid is true isn’t it reasonable to include it?

[One entertaining note on the 3% who said "yes": there is a suspicion from the IP addresses of respondents that they all came directly from trade union organisations!]

The problem for this survey, and all surveys, is how much information do you give people so that they can tell you their opinion?  ”Facts” are often not accepted universally as being factual.  But if you don’t provide relevant information people might make uniformed opinions that they would revise if they heard the information.

But if we ask “should they strike” and someone says “yes” and then add “should they strike if they are the highest paid” you get another problem.  People have just said “yes”, perhaps on principle of a workers right to strike, so some won’t change their answer when the new information arrives.

So perhaps the only logical solution is to ask different people different questions.  And to make sure that people only answer only one question so that they aren’t inadvertently primed by any preceding questions.

And what would you conclude from such a set of  surveys?  That people think different things depending on how much information they have and how you ask them about whatever it is.

Hmmm, not likely to make for a compelling headline in the press, or to give a company massive reassurance about a decision they’re considering.

And yet, despite these surveys producing such extreme responses that weren’t believed by the media commentators reporting them, will they keep going back to market research in a bid to understand what people think on a topic?  Of course, they will.

Why?  Because we are convinced we know what we think: we simply pose ourselves a question and find out.  And it seems entirely logical that other people must know what they think and that we can find that out by asking them.

But it just doesn’t work.

That’s the reason I wrote Consumer.ology.

So, to return to my road-crossing example, don’t be so sure that you would let the first hit by a car transform your beliefs.

Philip Graves

Philip Graves market research BA, banks, beliefs, British Airways cabin crew, market research, polls

Once upon a time there was a country.

They were having a very difficult time indeed.  Some greedy goblins had got out of control and now everyone in the land was troubled.  The money that the greedy goblins had taken had somehow disappeared, some said they had taken money that never existed, and now everyone was having to repay it whether it ever existed or not.

Everyone except the goblins that is, who had somehow found a way to get more money  for themselves – but that’s another  story.

Anyway, the country had a democratic process of sorts, of which it was most proud.  And in the midst of the “difficult time” it replaced one leader with another one, who thought things should be done a little differently.  This was done through something called “voting”, where those who wanted to shape what the country did explained their ideas and then people got to say which person they wanted to represent them when the decisions about what to do were made.

Some people didn’t vote.  Either because they had no interest in what was happening or because none of those asking for their votes said anything that impressed them – it’s hard to know which, but you could always “not vote” and some argued that it was the degree to which people “not voted” that really influenced the results.

But then there was a problem: the country also had other voting, not for the top job, but for the people who voted on what the leader proposed to do in the “difficult times”.  This voting was to make sure that the view of the people doing the voting was always current – it could keep the leader on track.  This seemed very fair.  And it would have been, were it not for the Opinion Pollsters.

The Opinion Pollsters were a fidgety bunch.  They didn’t like to wait for the voters to decide and hear the result.  They liked to find out in advance.

This was quite curious because the people doing the voting often changed their mind quite a lot, or else didn’t vote at all: in any event, very often what the people told the Pollsters turned out not to be the real result after all.

But that didn’t stop the Pollsters.

In fact, something very unfortunate happened.  The Pollsters started to influence what happened in the real voting.

To be fair, this wasn’t entirely the Pollsters’ fault.  Whilst it would have been nice if they had pointed out how hopelessly inaccurate their opinion polls were (especially when they were conducted quite a long time ahead of the real vote), they earned their money from conducting opinion polls, so that would have been asking a lot.

The Media Moguls were really calling the shots.  They too had seen how unreliable the opinion polls were, but that didn’t stop them running them as the headline story when what they showed reflected what they wanted them to say. 

When the people who thought they could represent other people well, heard from the Media Moguls’ polls that they were going to be crushed in the voting, they got scared.  They decided they didn’t want to be humiliated and began to talk about not competing for votes in the way they usually would.

They forgot about the importance of giving people a democratic choice; they were convinced by what the Media Moguls and Opinion Pollsters had told them that the people didn’t want them.  But they forgot not only that the opinion polls were often wrong, but also that things changed: a new attack on the country, a discovery that someone who wanted to represent the people was actually not a “nice person”, or a shift in the country’s financial fortunes, might all change how people felt and which way they voted.

But the Media Moguls and Opinion Pollsters didn’t worry about that: they liked the sort of voting that they could edit, shape and selectively publish that came from opinion polls, much  more than the kind that took place on a given date and with everyone taking part  (except those who didn’t want to) and everyone getting the results at the same time.

The Media Moguls understood that people could be like sheep and would follow the crowd if they were given enough reassurance that the crowd had decided something.  In other lands this approach to influencing the flock had caused all sorts of horrors – horrors that are far too unpleasant for this fairy tale.

And so the democracy about which the country was so proud  was killed.

But no one really noticed and they all lived happily everafter (especially the Media Moguls, Opinion Pollsters and, for some strange reason, the goblins).

Philip Graves

Philip Graves market research democracy, market research, opinion polls, philip graves