How Not To Involve Customers


In 1985 one of the biggest brands in the world nearly destroyed itself – by listening to what customers said.

Coca-Cola developed a product dubbed “New Coke” that was slightly sweeter than the original. Almost 200,000 blind taste tests were conducted and most participants said that they favoured New Coke over both the original formula and the companies bitter rival, Pepsi.

Hundreds of thousands of people can’t be wrong, right?

New Coke tanked – costing the company millions, with the CEO later commenting that they had “drawn a moustache on the Mona Lisa.”

There were two main lessons learned:

1: The research was flawed as it was based almost entirely on sip tests—a comparison of sips, not someone enjoying an entire drink. A blind test in a lab type environment was out of context compared to the experience of , say, drinking a Coke in the garden on a summer’s day

2: No-one realised the symbolic value and emotional involvement people had with the original Coke.  What customers said was “yes, this tastes a lot nicer” but when the product hit the market they behaved entirely differently. Influenced by their emotions and a status quo bias to keep things the same, they demanded the old Coke back.

Many of our organisations are still making these same mistakes – assembling focus groups and panels and involving users in ways that are wholly artificial compared to an actual customers lived experience.

In my last post I argued that we should know our customers, just never ask them what they want.  It drew quite a lot of comment and challenge with some points that merit further debate.

The top three charges were these:

1 – You’re saying professionals know best

No I’m not – but I do think the ‘customer knows best’ argument needs some challenge.

Whilst I agree that users are almost always closer to the problem than the average senior manager or executive, proximity to the problem doesn’t automatically make you the best person to solve it.

‘How can staff know the customer experience better than they do themselves?’ was a typical comment.

I think we are in danger of conflating two things here.  Experience of the problem and the creation of a potential solution.

As a road user I can sit in a traffic jam and moan about it all day long , but I can no more design you a solution than I can build a rocket to Mars.

Solutions require subject matter experts who have a deep knowledge of the root cause of the problem and can look through multiple lenses to craft a response.

The idea that a customer can provide that based on fairly limited knowledge (or interest) is naive.

2 – You’re saying customers are irrational

No – I’m saying we are all irrational and often make decisions accordingly.

At least three of my friends have chosen to holiday in the UK this summer citing concerns over terrorism. Statistically you’re 10 times more likely to die by falling down the stairs than you are in a terrorist attack – but none of them, as far as I know, are moving to bungalows.

Designing the right solution means testing for irrational behaviours.

And that can only come by observing what people actually do rather than basing decisions on what you think , or what they say, they’d do.

3 – You’re saying don’t listen to customers.

No – I’m saying just don’t let that be your only basis for decision making.

The research of Clayton Christensen found that companies that fail often listen to their customers too much.

As users we struggle to envisage the future. Indeed, asking someone to consider the future is a bit like saying, “Tell me how you will behave in five years time when I’ve rolled out the service I’ve just asked you about.”

At Bromford we are considering how we maximise customer input where it really adds value.

For us it’s the feed into the problem definition – and then at every stage through iterative testing, pilot and subsequent release (or shelving).

It’s balanced by robust internal insight measures as well. As this graphic by Tom Hartland shows – just because a customer likes something doesn’t mean it’s good business sense.

Screen Shot 2017-03-10 at 06.00.04


You’ll see that customers are generally not involved at the initial design stage – where we see that it’s our job to find out what they need, without directly asking them.

It’s useful to frame this in three steps:

  1. Know your customer. What are their likes and dislikes?
  2. Know your product. What are you offering and what problem does it solve for them?
  3. Know the thing. What is the ONE thing that the customer wants that makes what you’re providing special?

Saying “don’t ask customers what they want” isn’t anti-customer – it’s saying you need to work much harder at really knowing them.

Muhtar Kent, the outgoing CEO at Coca Cola said this recently:

“We’re still talking about New Coke and its lessons, which become clearer with each passing anniversary. We no longer control—if we ever did—the conversation about our brands. But we must be part of it. We listen, we analyse, we respond.”

Right now I’ll bet your organisation is developing services, transforming existing ones or getting hugely excited about new product launches – based upon what people say they want.

One of them could be your New Coke.

Question is, which one?


14 thoughts on “How Not To Involve Customers

  1. Great article Paul. This is also a good reason to keep things as small as possible. The smaller it is the easier it is to iterate. Also if you do have to throw it away (kill it) then the smaller it is the chances are it will not be that expensive.

    1. Thanks Barry – absolutely. By keeping things small and refining at that stage it should become easier to implement if it proceeds to project management. PM’s can concentrate on scaling and implementation rather than working with dodgy concepts!

  2. It’s an interesting idea – deciding what customers need without asking them (initially). It is a strange truth about the human condition that often what we want and what we need don’t match. Also, at times we don’t know what we want/need until we see it and either know that’s it, or know that’s not it. Is your stuff based on the Agile approach?

    1. Thanks Lyn – we do use elements of agile. Particularly breaking down bigger problems into smaller units and then iterating on them. We try not to be limited by one particular methodology at design stage. It’s a different story for evaluation , or project management obviously where more structure is required.

  3. For me digging into why people say what they want is the key thing.

    What needs to change? Rather than, what do you say you want as a solution?

    Asking people who’ve never had a job or are just finishing education what they want to do as a career is my favourite example…it’s completely bonkers…like asking me what kind of opera I’d like to write….I have no idea where to start or what I do/don’t value in an Opera. But finding out what motivates me and what I enjoy doing…that can tell you as an expert narrow in on the kind of experiments you could involve me in to help me decide what opera I like.

    A big issue I have is that client engagement isn’t a one shot process, you have to go back and check as well, people’s lives change all the time, offers based upon a narrow snapshot of questions asked two years ago might not work now…

    1. Thanks Niall – the job example is perfect.

      When we did some work with the long term unemployed we found a lot of them had only ever been asked what they ‘wanted to do’ (a very vague nebulous idea) rather than to choose a career or start a job search based on specific interests. By focussing in on something they enjoyed doing they were able to narrow their frame of reference and start building something.

      Great point about one shot process as well. Thanks for commenting

  4. Think to some extent you have the wrong headline here?
    Could argue that New Coke failed because the research was a bit crap…surely a reason to get the research right, not to stop doing any?

    1. Well, I think it was flawed rather than crap. It’s used to illustrate that people don’t act as they say – and yet what customers say is what many organisations base decisions on.

      In no way is the post anti-research. Far from it

  5. It’s really good case study, thanks for the post!
    I think that one can get to know great insights from the customer about the PAST and PRESENT, and a lot of misleading opinions and assumptions about the FUTURE.
    Nevertheless, I find it quite easy to fall into the trap of basing decisions on what we or the customer thinks they would do in the future, particularly because it’s so often feels like a much faster shortcut to just take the customer’s answer about a theoretical question than to properly observe and understand the current behaviours. But as usual, shortcuts don’t pay off on the long run 🙂

    1. Top comment Robert and I like your honesty. I too have fallen and still do fall into that trap as it feels faster, more intuitive and – let’s face it – easier than proper observation!

  6. Can you ever ‘maximise’ anything?

    When you believe you have you stop! For example fine tuning a car engine to get 135 mph out of it is not the maximum or be perceived that way. There is always a way to get 136 mph and then 137 mph.

    Then you appear to want to stop this constant search to actually maximise if that can ever be done and on the basis of what YOU perceive to be added value from the customer and thus apply another constraint.

    Constant improvement is much better than this finite perceived maximisation which in all reality is a self-imposed constraint on improvement?

    Additionally is the clever graphic above any different to the 4 “C’s” of Best Value from all those years ago which is an iterative cycle of constant improvement and NOT accepting that maximisation exists? I suggest it is.

    Further so many external variables outside your control change within housing which makes any notion of having achieved a maximum nonsense as what you believed to be that maximum constantly changes too as it operates in a constantly changing context.

    PS New Coke failed because it tasted like shite? Yes!

    1. Really the word maximise was poorly used – it could have easily been ‘deploy’ or ‘use’.

      Essentially the post is about how easy it is to fall into the trap of basing decisions on what we or the customer thinks they would do in the future.

      That was the New Coke lesson.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s