42 Degrees - Business from a different angle

It doesn't matter whether you are starting a new business, launching a new product or simply extending an existing line, you should not risk a penny before you have asked your customers. Read almost any business textbook and that is the advice you'll receive.

And, of course, it's true. It does make obvious sense to ask people what they want before you give it to them. Up to a point. But when you look at many of the most successful and powerful businesses in the world, you'll find they do the exact opposite.

Did inventor James Dyson carry out focus groups before launching the revolutionary vacuum cleaner that bears his name and which made him an £800m fortune? Did Bill Gates consult quantitative surveys before launching the Word operating system that was to make him the richest man in the world?

In entertainment, did Mick Jagger ask for his listeners' preference ratings when choosing the songs that were to earn him £85m in live performances alone last year? And in an old-fashioned product area such as food, did Marks and Spencer rely on hall tests before launching the chocolate pudding that was its bestseller last year?

The answer to all these questions is, of course, "no". Which raises another question: what is the point at which you have to stop listening to your customers and start dictating to them?

The problem is that, just as there is clearly no single answer to the question of whether or not you should market research your new idea, there is no single answer as to why or when you should not market research your new idea at all. Each of the examples above has different reasons for relying on gut instinct. They suggest that the case for market research is not always quite as clear-cut as it seems.

James Dyson's big idea was to produce a bagless vacuum cleaner that would not clog up and lose suction power. To do this, he famously experimented with cyclone technology and produced 5,127 prototypes. But it only worked properly when he introduced two cyclones. The problem is that customers are not experts.

"If you had asked whether they wanted dual cyclone technology in their vacuum cleaner, the average person wouldn't have a clue," points out Sanjay Nazerali, chief executive of market research firm, The Depot. "If you simply asked whether they wanted a better vacuum cleaner, people would say yes. But Dyson probably already knew that,"

When Dyson first launched his cyclone cleaner in Japan, he had in mind a premium product. He could not have conceived, however, that it would sell for the equivalent of £1,200, at least six times the price of other cleaners. Similarly, in 1943 Thomas Watson, chairman of future computer giant IBM, famously said: "I think there is a world market for maybe five computers."

"No great innovation was ever a certainty," says Nazerali. "Even massive successes such as MTV or mobile phones were risks because customers just don't know how they might behave in the future, especially in response to something that doesn't currently exist."

The suggestion that you might not research something because you are in a hurry sounds as if it could be a justification for sloppy and impetuous decision-making. But even a couple of decades ago, when markets were relatively static, companies might market research new products for two or even three years.

Today the rate of change is much faster and many businesses are like the fashion industry: opportunities open up and disappear within 18 months. As advertising magnate Sir Martin Sorrell is fond of saying, "It's better to be wrong on Monday than right on Friday."

His point is that sometimes it's better to be fast to market and then change your offer than make the perfect product when the opportunity has vanished. Hence the rise of the so-called "soft launch", in which products are eased into their market place without fanfare in order to iron out any wrinkles.

Not only can lack of sufficient time be a reason for bypassing research, so can lack of sufficient funds. Do banks research every new nuance of their new offerings? Does Marks and Spencer research every new sandwich? No, of course not. Formal research can be very expensive. "If the upfront investment required is low you might be better off live testing your new product in a region or key stores," says Nazerali. "If it works roll it out. If not withdraw it,"

This is another reason for the growing popularity of so-called "soft launches". They are used not only when products need low up-front investment but also for those with complex consumer interactions, such as restaurants and websites.

In life and business there is a direct relationship between risk and return. One of the reasons for formal market research is to reduce the risk of failure. But just as it reduces the 'downside', it can also reduce the 'upside'. If you want a step change in margins or sales, too much consumer research can get in your way.

Soap giant Procter and Gamble acknowledged this recently when it admitted that its famous reliance on market research was producing solid but uninspiring growth. It now needed "double digit" growth and how was it going to achieve this? It was going to stop looking to customers for ideas for new products and turn to its boffins. "Certainty is a killer of innovation," says Nazerali. "If you want certainty, you'll tend to develop the same old products in new colours and flavours. But if you want something really new, with real growth potential, your customers are not necessarily the place to start."

Now only a fool would take these arguments and conclude that there is no need for market research, ever. You always need to understand your customers and you can only do that by talking to them, say experts. "You need to understand where your current products fit into their lives and you need to be very aware of tensions in their use," says Peter Shaw, managing director of marketing consultancy Brand Catalyst. "The important thing is not to take responses literally. You have to use information diagnostically, not just as a tick or a cross for your idea."

So, like the drunk and the lamppost, you should use market research for illumination not support. But you also need to be aware that there are some areas on which it cannot cast sufficient light to be useful.