To Innovate, You Have To Stop Being A Slave To Data

By their very nature, people are resistant to change. So if your goal is to innovate, why would you listen to them?

We live in a quantitative, data-driven world. Design, product development, and marketing decisions are all based on careful analysis of data that have been gathered through sales, market surveys, clicks, eyeballs, and focus groups. At the same time, there’s a clarion call for brands to innovate and differentiate.

It is well established that people resist change and don’t like being different. Yet brands are asking these people, who don’t even care for change, how they should change. It’s like asking a priest if atheism is a good choice. This doesn’t mean that brands should stop innovating and stay stagnant; that would lead to a slow death. It does mean that in many cases brands, and especially creative brand communications initiatives, overreact to what the data "is telling them." They are often asking the wrong questions and misinterpreting the meaning and significance of the data.

The simple fact is that it takes time for people to accept change, and if asked on the spot for their first reaction, it is likely to be negative. Now there’s nothing wrong with using statistical analyses and focus groups. In fact, they are extremely useful, especially for studying behavioral traits and figuring out what people need and want. The problems arise when a new concept is introduced to the sample or focus group. It’s perfectly valid to ask, "What do you associate with this brand?" but pretty counter-productive to ask, "What do you think of this brand refresh?"

Change Is Good

People have a propensity to choose what they are used to because it makes them feel safe; it is recognizable. A lot of people get freaked out at the prospect of some brand agency changing the look of their morning O.J.: How will they be able to find it in the fridge?

The truth is that people are exceptionally good at adapting to change, and if done in the right way, it can be tremendously beneficial. We are, after all, the products of evolution. Approximately 100 years ago, Frederick Taylor, the father of scientific management, studied production rates in factories and discovered that small changes in the environment and processes increased productivity. For example, if you switched on the light, people tended to work a little more efficiently, and if you then switched it off again, instead of going back to the previous productivity levels, efficiency would actually rise a bit more. So incremental change seems to be good for ROI.

Knee Jerks

Unfortunately, too many brands have a tendency to treat market data as law. So much so that it will override the expertise, experience, and intuition of designers, strategists, and other creative professionals. The result is usually something bland, traditional, neutral, and ultimately void of meaning and voice as if designed by committee.

Social media has made this process painfully visible and is now looked upon as a focus group in and of itself. We have seen brands publicly recoil in fear as soon as a group of online sourpusses proclaim their distaste. The "consumer," as represented by surprisingly design-conscious, self-promoting bloggers, doesn’t like the new logo, so the company gets second thoughts and surrenders.

Facebook and The Office

People do eventually adapt to novel introductions, even after their strong negative reactions. Every time Facebook introduces a change in its user interface tons of people rally together in groups and protest loudly, proclaiming how awful it is and how inconvenienced their lives are. When the next update comes along, those same people are again complaining and saying that the current version, which they previously hated, is fine.

Likewise, when the BBC tested the groundbreaking comedy series The Office on audience focus groups, it apparently tested among the worst in the network’s history. The network decided to stick to its guns, and lo and behold, now it is a global, multimillion-dollar asset. People didn’t like it at first, because it was different from what they expected a comedy show should be. That is often the case with disruptive, game-changing products and services.

Change is necessary in order to stay relevant. Stagnancy leads to obsolescence. Brands need to be brave, bold, and confident in their intuition—and stop being slaves to data that doesn’t carry real meaning.

[Images: Blaj Gabriel and Mark F Gutierrez via Shutterstock]

Add New Comment


  • DASH

    something that I have been thinking on since last year...extremely well conveyed. In India this is an eternal war topic between designers and managers across organizations. Another addition to the above anecdote is : "Lets go with the standard." A common dialogue from the management to end conversations or play safe. A simple question, What is 'standard' and how much relevance does it hold today. Is the statement even valid, so to say.

  • ken nohe

    Yes but by being to far ahead of the curve and "sticking to your gun" you can easily go under before your ideas are finally accepted. 

  • Storsdors

    An interesting take on Clayton Christensen's idea of Disruptive Innovations.

  • Ari Nave

     Well, work has been done to show that the underlying premise - that people are not open to change - is not quite right on 2 fronts.  First, it seems that the population exists has a continuum between traditionalists and experimenters.  See for instance Jonathan Haidt's talk on the roots of liberalism/conservatism. 

    In addition, evolutionary psychology has a great deal to say about the issue.  As being adapted to leverage culture as a tool of survival, our appetite for change in in relationship to the amount of change in the environment.  The argument goes like this:  in an environment that is very stable, the masses are probably following cultural variants that are not too bad, so to deviate from them and engage in individual experiments is probably very costly with little payoff.  On the other hand, when the environment starts to fluctuate greatly, the chances of learned behavior remaining adaptive drop and at some point you are better off experimenting to find new behaviors that are better adapted. 

    On a population level, you would expect there to be an approximate Nash equilibrium where the proportion of experimenters to traditionalists settles in relationship to the degree of environmental change. 

    Finally, folks like Robert Sapolsky have strong evidence that we become less open to change as we age.  Makes sense if what we have done has worked for us and kept us alive for a few decades.  So we get stuck in our ways. 

    So if you recruit for young open-minded experimenters, you are likely to have an group of innovative thinkers to engage in participatory work...  if you craft an environment that engenders such performances. 

  • Philip Graves

    Yes, there are a couple of misjudged conceptual associations in the article, but the fundamental thrust of it is absolutely correct.  

    The reason that people can't react in a meaningful way to new ideas in MOST market research settings is three-fold. Firstly, the actual consumer context, which is crucial to determining behavioural response, is typically absent.  Secondly, the research is USUALLY interrogating the conscious part of the mind, when most behaviour is driven by the unconscious mind.  Thirdly, the process of soliciting or obtaining responses is, in itself, a source of bias (and not one that can be removed by questionnaire design or expertise in moderation).

    The above is based on science not opinion.

    Are there solutions to this?  Yes and no.  The role of research has to be entirely reappraised and, with the best will in the world, people who have convinced themselves that asking consumers questions is a reliable and valid route to understanding consumers, are poorly placed to distinguish 'right' from 'wrong', methodologically speaking.  

    Live testing is required: if market researchers can manage this with objectivity then that's great; but if they feel that soliciting consumer opinion is a useful reference in the analysis mix they will entirely negate the benefits of experimentation.  

    Anything that attempts to predict the future on the basis of something other than past behaviour should be consigned in the same bin as astrology, tea leaf reading and all manner of related mumbo jumbo.  Anything that attempts to predict the future on the basis of existing behaviour should accept it's probably right, but it's got very little chance or distinguishing a good innovation from a bad one.  And once the innovation arrives the predictions need to be re-wriiten.

    Understanding HOW customers think is, generally speaking, much more useful as a foundation for innovation than trying to understand WHAT they will think.

    A government decided to innovate communication in the energy sector to encourage people to turn down their heating.  Someone suggested that telling people how much money they could save by turning down their thermostat by one degree would be a great way to create the change desired.   More than $100k of research later, and this self-evidently sensible idea had been validated.

    However, when an experiment was performed by a behavioural economist and groups were sent one of three different messages about energy use, the results were based not on what people said in research but on their meter readings. The results weren't quite as the research had suggested: people being told how much they could save by turning down their thermostat one degree used MORE energy, not less. 

  • Simone Oltolina

    Mmm, this is a very delicate topic and I'm afraid the article is a bit simplistic. Innovation in in fact a 2-step process: first you have to generate some ideas (and you can't rely on the consumer for that, although consumer knowledge does fuel the thinking), then you have to test those ideas and that's when the 'quantitative'/data-driven approach comes into the picture.

    Now, first of all it must be said that a lot of this 'testing' stuff ends up killing ideas which might have been promising either because the research design is flawed or the  respondents are poorly recruited.

    Research design  is obviously key. Asking 'do you like this' won't give you much in terms of insights, you have to be more subtle and understand basically if your ideas conveys what they are meant to convey. If that happens, you can accept a mild level of rejection, especially if you know where that comes from.

    I'm all for creatives/strategists and visionaries, they have a large role within any innovation process but that's about it, you can't just rely on them to take a decision.

  • tomGlimps

    Thanks! that was a great read. Good point You've touched here & wise words came out! There always be some who would rather talk and argue, instead of adapting or start learning fast/early enough..but after a while, they simply turn over & go forward untill the other obsticle appears - that's the way of living for some.

    This article will spark a fire..& that's good thing.
    keep it up!

    Thanks once more for great read!


  • Patrick

    In the segment titled 'Change is Good' you argue that 'incremental change seems to be good for ROI' on the basis that changes in lighting having a positive effect on work place efficiency. The reality is that the experiment being referenced caused an increase in productivity because the subjects of the experiment knew they were being watched and worked harder after each change because their output was being scrutinized. This is now know as the Hawthorne Effect. 

  • AmandaS



    I don’t entirely disagree, and I've been an analyst for over a decade. It always depends upon the questions you are asking of the data and knowing the
    limitations of that question and it’s answer. You always have to make a guess
    at how markets will react to a change in branding or comms. You can only make
    fair assumptions on that reaction based on data that you have, what you know
    about people prior to that change, or how consumers have reacted in similar
    situations. No one can predict the future – but the intel we hold on our
    consumers, competitors and markets are the closest thing we have to a crystal
    ball. Mystic Meg doesn’t always get it right ;)

  • Peter Norman

    What rubbish. To summarise: "All users are idiots and designers know best". This kind of thinking was out of date 10 year ago and there's really no excuse for it today. Focus groups don't work because they focus on opinions, not behaviour. You're ignoring the behaviour-based data that you get from usability tests, web analytics and contextual inquiry. It's these kinds of research method that fill in design teams' blind spots --- and helps make truly great products.

  • @robertotra

    Uhm... basing the definition of this article as rubbish upon the goodness of behavior tests is (in my opinion) missing the point: usually you analyze behaviors in order to evaluate, confirm or refine a strategy or a product, but the core ideas of the product and strategy are already formed. Of course we cannot rely only upon our intuition or subjectivity and need confirmation from external reality, but is also true that we cannot invent anything new if we rely only upon external data. So, as a mix of opinion and behavior, we need both to make good stuff. But it has to be underlined that if a good new idea may eventually arise from intuition without data, no new idea can come from data without any intuition.