Four customer insight traps that could cripple your business

At the heart of business is the need to understand and influence people. Sadly, the emphasis is usually on the latter with a somewhat fleeting attention to understanding. Absurd really, when you think about what is at stake. At the heart of any good strategy is a key insight about the future, based on data gathered in the present, alongside what is usually a significant investment in time and money. So, you’d think that professionals would attach reasonable value and importance to these key insights and approach them with due consideration and forethought. More often than not, however, the intensity of the short-term business demands proves too distracting. To that end here are four traps, with responses to each that will help you to develop a rich foundation of customer understanding from which your business can flourish.

Ask enough people what they think, and with any luck, some of them will tell you, but that doesn’t make it a true statement about the world, but simply what people think about the world, or worse what they falsely believe they think. We are all victims of a multitude of different unconscious biases, (some of us more so than others), so the belief that asking questions will secure a true representation of the world is borderline delusional. Far better to observe what people actually do, rather than ask them what they think. Yet, we find ourselves in a world obsessed with market research, with businesses spending millions on brand surveys, then working to conclusions formed through questions asked years beforehand. Actually, I have some sympathy for this position. The pressures on marketeers to be answerable and measurable invariably leads to the ‘terror of the tangible’ (a drive towards valuing only that which can be measured), but not necessarily to the truth. Equally, behavioural studies are, by their very nature harder to define and execute, with data that is less qualitative and thus, harder to swallow for an increasingly binary society. There’s also admittedly something immensely gratifying about asking a question with multiple choice answers, and being able to say, with ill-conceived certainty, that x% of people think y, and consequently don’t think z. 

the belief that asking questions will secure a true representation of the world is borderline delusional

However gratifying it may be, most surveys fall foul of being generated completely out of context. The reality is, people are fickle. We may feel one way, one moment, in one set of circumstances, but with the slightest of changes, feel distinctly different in another, whilst remaining utterly unaware of any change. How we think when we’re sitting at our computer, answering a bunch of questions about what we might do, late in the evening, our motivations laced with the distinct hope of winning a suitably chosen, (but supposedly non-influencing) gift, sipping a cold beer with one ear aligned to a comedy podcast in the background; is all a far cry from what we were actually thinking at the point we made a purchase decision. Time, environment, and a whole host of other factors that create ‘context’, have an immeasurable influence on how we think. Some factors we are aware of. Most, we are not.

With a little ingenuity and consideration, it’s almost always possible to establish a series of behavioural tests that study audiences in the moment, completely in context. It’s usually a lot more work, and the results are sometimes less binary, but far better to gain insight around what people actually do, than be misled by what they think they might do, or have done in the past. Even in the circumstances where you’re trying to test a concept that doesn’t yet exist, far better to create gorilla-style real life experiences that can tell you about how people are actually motivated, rather than asking them questions completely out of context. So, when you’re looking to understand people, always try to find a way to study how they behave, rather than asking them what they think.

The difficulty in generating good behavioural insights leads me to trap two; the snapshot trap. 

Culture is never static. The world itself, and how we think about it, is changing all the time. The idea that an insight gained at one point in time is going to be true a year, or even six months down the line is, for this very reason, fundamentally flawed. Too often, business decisions are based on a survey carried out a year or two earlier. Not only have they made the mistake of asking people what they think, but they continue to rely on the same false fact a year later, because the act of reaching out to ask the same target audience how they feel now, is either too expensive, too intrusive or too much work. The opportunity cost (of triggering a host of badly chosen actions based on old data and new assumptions) usually dwarfs the time and money required for fresh insight, yet this doesn’t seem to perturb many - such is the hunger and impatience for growth. That behavioural research is usually that much harder to execute only exacerbates the situation. The harder it is, the more likely an organisation is to rely on old data, again failing to measure up the cost of doing so against the larger investment that will be made in its name.

The answer is to avoid a snapshot research mentality entirely, and to stop treating customer understanding as a peripheral activity and a series of one-offs. Make the study of consumer behaviour a central pillar of your ongoing business operations. One only has to look at the tech giants of today to see this modus operandi in full effect. But I have one big caveat to add. When your creation of tools to develop customer insights is driven by profit alone, rather than serving the needs of your customer, you run the risk, as most of the major tech behemoths are experiencing right now, of swiftly eroding customer trust. One hopes that the successful corporations of the future will lead with a mantra of ‘with great customer insights, comes great responsibility’. Perhaps not. Time will tell, but consumers are increasingly sentient of being made to feel like a commodity, so ignore their sensitivities at your peril.

So, you’ve switched to a focus on behaviour in context and made the study of that behaviour central to your business. You’re now ahead of the game but please don’t let it go to your head. A modicum of humility is required. Behind every good behavioural test, there is a hypothesis, and inevitably a degree of interpretation. For visionary entrepreneurs with grit and an appetite for success, it becomes immensely hard to divorce themselves from their hypothesis. Confirmation bias gets the better of them, and somehow they manage to see proof for the hypothesis, no matter the data. It seems unreasonable for me to say, ‘drop the ego’. None of us are truly without ego, and all of us are attached, to varying degrees, to the hypotheses we have about the world around us. Better to surround ourselves with people who make it their business to be sceptical. Humility is rare in entrepreneurs, but where once the more visionary individuals were encouraged to pair with accountancy-minded counterparts (watch that bottom line!), the age of the data scientist has now arrived. The scientific ethos - of applying careful observation and rigorous scepticism about what is observed - is an absolute must if you are to reap the benefits of good research. With all that work going into gaining customer insights, are you able to accept when the data tells you that your hypothesis is wrong? If no one on your board can boast scientific humility and scepticism, then you’re highly likely to get carried away with yourselves and fall foul of confirmation bias. So, thirdly, remain sceptical and open to being proved wrong. There’s value in the lesson, if you’re willing to receive it.

Finally, we see the world through our own lens. To that end, some things fall into focus, whilst others remain in the periphery (with many of course falling into blind spots). So it is with customer understanding; a single lens will not do, and especially if that lens offers a data point for how people think, captured only at one moment in time and distorted by confirmation bias.

Indeed, this even extends to the often lauded maxim of Net Promoter Score (for the uninitiated, NPS is a widely used market research metric that typically takes the form of a single survey question asking respondents to rate the likelihood that they would recommend a company, product, or a service to a friend or colleague). Data has indeed shown that high NPS often correlates with profitability, but when you consider how that data is generated, you immediately see that even the magical NPS fails to bypass bias and is asked outside of the true context - the exact moment that someone actually recommends a product. May I remind you that people are fickle.

when you consider how that data is generated, you immediately see that even the magical NPS fails to bypass bias and is asked outside of the true context

Now, I’m not saying drop the NPS scores. People understand them, they’re easy to present, and they still seem to hold some weight in most business circles. (Please don’t present them on their own though, without a comparison NPS from your market. There is no such thing as ‘a good NPS score’ in isolation.) No, what I’m advocating is more lenses. There are so many alternatives to NPS, such as repurchase rates, time taken between purchase, social mentions and other forms of advocacy, that measure behaviour rather than thoughts, bypassing biases and resulting directly from the context you want to study. Measure these too, and automate the process. 

One channel of data can mislead you. Two lines of data and confirmation bias can still get the better of you. But when you manage to remain sceptical and observe the customer through multiple lenses, especially lenses that view behaviour over time, then and only then does the world suddenly become a whole lot clearer.