When Samuel Goldwyn wryly advised that you should "never make predictions, especially about the future," he may have been reflecting on Harry Warner's misplaced attachment to silent movies, who in 1927 asked "Who the hell wants to hear actors talk?".

The perils of prediction are legend and the source of many chuckles, but making bets on the future is part of a designer's job description. Whether selecting a colour, or devising a product strategy, designers are--consciously or not--making a series of assumptions about the future. So how can we shorten our odds and become more prescient?

The first hurdle to clear is the cynicism towards looking at the future that Goldwyn's quote captures. The cause of analysing change in a systematic way has not been helped by the self-importance of so many forecasters. Indeed James Woudhuysen, Professor of Forecasting and Innovation asserts that "Because of today's exaggerated sense of uncertainty and foreboding about the future, forecasters have never been in greater demand--and have never been less credible!" Whether they be economists making predictions to three decimal places based on bogus computer models; black-clad Parisians foretelling the return of the 'New Black'; or intrepid coolhunters reporting back from their latest Tokyo shopping expedition, they often have the whiff of snake oil.

The future is much more important than the new, the hip and the cool. The fashion and communications industries may rely on a constant feed of zeitgeisty whims, but product designers and strategists need to develop a deeper understanding of change. Our timeframes are too long and problems too complex for such fluff. Management thinker Gary Hamel frames foresight as being "prescient about the size and shape of tomorrow's opportunities" by building an "assumption base about the future" based on "deep insights into the trends." His advice that "strategy must be created from the future backwards, not the present forwards" also underlines the short-termism of much of today's innovation activity which places too much emphasis on what consumers said last month. When done well foresight can help designers make sense of a world in flux, bring clarity to planning, and help situate strategy within a future context in a way that can be communicated to senior management.

Eight tips to shorten your odds It is now 10 years since I did my first foresight project. Its horizon was the world of 2010, which is now all but upon us in product planning terms. As I write I am now midway through another look forward for the same client, which has been cause for reflection of what I have learned along the way. Here are eight good habits I learned to adopt:

1. Frame the scope and focus Too many foresight projects suffer from lack of specific structure and focus. A general trends presentation is wheeled out for all occasions too often. If we define trends as useful simplifications, abstractions or generalisations of notable shifts in reality, we need to be clear on what part of this reality we are interested in and why. For example, we have been asked by corporate strategy departments to help them scope what business areas they should be in 10 years hence, and design departments wanting to develop a new design direction that will be resonant and distinctive for the next couple of years. One is a question of long-wave social, economic and technological trends; the other focuses on shifts in particular consumer segment preferences, industry sector constraints and relevant design trends.

Once we have a clear idea of scope and objectives, the next question is how to decide on which trends are relevant. The trick here is to develop an explicit set of filtering criteria that helps the project team to make calls on which trends are likely to influence the reality in question. There is no systematic way of developing these criteria, but developing some working hypotheses early and then refining them en route through discussion helps to maintain focus.

Once we have framed the space, the next challenge is to break it down in a useful way by constructing a framework that structures the trends to illuminate their hierarchies and inter-relationships. For example, it is useful to understand that some trends are macro-drivers of smaller sub-trends.

2. Take a long and wide view To get beyond the hype, fads and received wisdom of a particular time and place, it is vital to develop a big picture perspective. Foresight work is about identifying and understanding discontinuities, which are best spotted from 35,000 feet.

It is often sobering to situate a trend in its historical context. Taking a long view provides more dots to join and a better sense of rate of change; as forecasts often get the phenomena right but overshoot or undershoot on the timeframe. However, as Larry Keeley has pointed out, "we always overestimate the amount of change in the short term and we underestimate the amount of change in the long term."

A classic example of this pattern of events is the adoption of 3G mobile technology. Back in 2000 the UK government held an auction for 3G licenses, and the "winning" mobile carriers handed over $40 billion for the privilege of being able to offer 'fast' mobile internet services. After much initial excitement among technology investors and journalist about the pocket internet, 3G's development was plagued technical problems and an underwhelmed public, until a few years later the carriers wrote off much of their investment as an ill-advised gamble.

As I argued in 2001, 'The hype around the mobile Internet always suffered from too much business and technology push, and not enough user pull.' Eight years after that auction, the 3G iPhone finally delivers on the promise of the mobile internet experience. The techn-optimists and pessimists are usually both wrong.

The hype cycle is a useful way of visualising this evolution of some technologies from peak of inflated expectation, through the trough of disillusionment to the plateau of productivity.

Another lesson from the past is how the future is always a mix of revolutionary change and evolutionary continuity--and sometimes regressions are in there too. Unlike the Star Trek view of the consistent future, many of today's trend-setters enjoy gaming on iPhones and organic gardening. History does not follow straight lines.

As well as history, an important key to the big picture is cross-disciplinary bandwidth--the ability to understand an issue across a range of domains. Fitting the wide-angle lens helps reduce the likelihood of being 'blind-sided' by a key development in another area. Collecting data from a range of areas that interlock around a particular issue also gives us greater confidence of its importance. For example, we particularly value the insights from our network of sociologists around Europe.

This ability to pan out for the detail and use interchangeable telephoto and wide-angle lenses has to be developed over time, but constructing a team that contains a few grey hairs and a mix of disciplinary backgrounds is a great start.

3. Put sociology before technology Whenever people look to the future, the first place they tend to turn for guidance is technology. Now that walking on the Moon has more to do with history than with the future, it should be clear that the future is not purely driven by technology. In the past 50 years we have supposedly lived through the Atomic Age, Space Age, the Supersonic Age and more recently the Information Age of the paperless office and the end of the book.

Two of the most notable shifts in Western societies over the past few decades have been the elevated role of women in society and the fragmentation of societies into a culture of mistrust and fear. Neither was widely predicted; both have little to do with technology.

Sure technology shapes us, but we also shape it, and looking at the future through the prism of a technology is bound to lead to a blinkered and distorted view.

4. Get under the surface Once is an accident, twice is a coincidence and three times is a trend. Foresight demands more than the pattern recognition of a trend spotter. To understand change, we must get beyond mere description of surface effects, and also beyond simply extrapolating existing trends into the future. If we do not understand why something is happening, we cannot anticipate how important it will become, potential tipping points, or how long it will last. To really grasp the essence of change, you have to get to grips with the dynamic nature of trends--how they have developed, and their interactions with other trends as both drivers and inhibitors.

I will never forget the lunch I had with a pension advisor back in 1999, when the tendency was still towards earlier retirement. The 2010 foresight project I had just finished had convinced me that this trend would begin to reverse and when I told her that I expected to retire at 70 she nearly choked and fell off her chair!

5. Be an informed contrarian Most forecasts are wrong in the same ways, because visions of tomorrow tend to be projections of prevailing hopes, fears and prejudices. It is always wise to be a well-informed skeptic and challenge the widespread orthodoxies of the moment. Be critical, ask hard questions and search out countertrends. Trust your gut and then look for supporting evidence.

One of the mistakes I made back in 1999 was to accept the received wisdom that GenXers were looking to friends rather than families as support networks--a perception reinforced by shows like Friends. If we had looked harder for data, we would have found that they were spending less time with family but even less time with friends.

One we got right was to challenge the idea that a decade after the end of the Cold War, belief in the free market had triumphed. We talked rather clumsily about an emerging 'search for authenticity as a reaction to the marketisation of everyday life'. The Seattle riots kicked-off the week of the final presentation of the project and Naomi Klein's 'No Logo' was published the following month.

6. Have a point of view A presentation of 20 trends does not add up to a view of the future. Many trends contradict each other, and choices have to be made. For example, the shift towards ethical (premium) consumerism and the economic crunch are not compatible; something has got to give.

In design one visual trend no longer dominates. Today minimalism co-exists with the baroque. The question is which trend if any has any relevance to the brand or product in question.

Your client or manager wants to know what it all means, and how the contradictions will resolve themselves in their specific business situation. Foresight is not really about making predictions, but presenting a credible, coherent and compelling narrative about the future as a stimulus for discussion and decision making.

7. Be specific about impacts Many important trends are irrelevant to particular companies and projects. I have lost count of the number of times I have sat through a presentation of niche über luxury trends when the client is firmly in the über mass market. It is always wise to think through the relevance of particular trends to particular industry sectors, product categories or market segments. In many cases, a nuanced analysis of an established trend's impact on a company, category or project is far more useful than breaking news on the latest fad.

8. Track and tack Foresight is best treated as an iterative exercise and not a one-off exercise. Futurist Paul Saffo advises forecasters to "hold strong opinions weakly... If you must forecast then forecast often--and be the first to prove yourself wrong." We build a framework of assumptions about the future, and then review and update it on a regular basis. It is an incredibly enlightening exercise!

After recommending a visual design language identity to one client, which was partly based on trend analysis, the client came back to us the following year to ask what was next. We explained that not only was the direction still right for their brand and product category, but was aligned with a long-wave trend that would give their visual positioning resonance and consistency for several years. Instead of jumping on the next trend, we helped them evolve their visual identity for a number of years, before recommending a new departure when the time was ripe.

So Goldwyn was half right, we cannot predict the future--illustrated by one of his own tongue-in-cheek predictions that "Flashbacks are a thing of the past." But that should not stop us thinking systematically about the future as it readies us to shape it more effectively. After all, Goldwyn quit the rag trade for the fledging movie business, founding MGM and discovering Gary Cooper in the process. Not a bad bet in retrospect.

Kevin is the founder of Plan, a product strategy consultancy based in London.