Do You Believe in Oracles? Trend Forecasting is Trending in the Fashion Industry

In March, I got a special edition of The Economist in collaboration with Le Monde, which brought articles focused on the year 2020. At the beginning of the publication, there were some predictions of how economy and politics would be this year and by reading it, it was clear this was written, approved, and printed before coronavirus was a thing. The predictions were so insanely wrong that it got to the point of being funny to read them. Of course, the pandemic had a quite unexpected impact, but this got me thinking.

The fashion industry loves forecasts, and with the Covid-19 impacts, they have never been trendier. The fashion industry was one of the most affected by the pandemic and businesses, insecure about the future, are looking up to trend forecasters for help. Platforms such as WGSN are extremely successful for doing almost solely trend predictions. Don’t get me wrong, I absolutely love reading these platforms, but I am a bit sceptical about trend forecasting. Some predictions do carry in-depth research and a realistic outlook. Others are closer to clickbait sensationalist content. The thing is, both of these very often end up being wrong about the future. When they are right, is not always because they predicted something. Trend forecasting is massively used not only in the fashion industry, but also across all types of businesses, with the aim of helping professionals to make important decisions. So, I can’t help questioning: how accurate are trend forecasts and how much should we base important decisions on them? The answer is way more complex than a yes or no, and I will try to explain why.

The Basics About Trend Forecasting

First of all, if you have no idea what trend forecasting is, especially in the fashion industry, let me quickly present it to you. In a few words, trend forecasting is the technique of predicting the future direction of something, such as buying habits and moods of consumers. In fashion, it also includes anticipating the fashion trends that major designers or influential individuals will offer. From identifying the future colours and fabrics that will be on runways to which social media consumers will prefer to use next, trend forecasting is broadly used in fashion businesses to make strategic decisions with data validation.

These trends can be broken into micro and macro trends. Microtrends are those with a smaller lifespan, such as bell sleeves and gladiator sandals, for example. Macro trends are movements that have a stronger impact on the industry, as the rise of sustainable fashion or the boost of social media. Trend forecasting is often presented to consumers in fashion magazines, social media or websites, but I rather focus on the B2B and “analytical” trend forecasting, done by specialized platforms or consulting companies. These are the ones businesses often use in their strategic decision-making.

Talking about trend forecasting and not mentioning WGSN is almost a sin in the fashion industry. The platform offers fashion predictions of up to 2 years in advance. Forecasting colours and trends to companies that sign to their content, their subscription is not cheap, going up to the tens of thousands of dollars per year. But there are many other platforms that tack into trend forecasting in the fashion industry (even if not exclusively) such as Mintel, Fashion Snoops, Pantone and Pattern People. Consulting companies also explore this market. McKinsey, for example, produces an annual report called The State of Fashion with the main trends for the year coming.

Is Trend Forecasting an Art, a Science, or Both?

To answer this question is first important to understand how trend forecasting is done. To predict the future of the fashion industry, forecasters use many “tools”, I will focus on three: big data, research and — here lays the subjectivity — intuition.

In the last years, big data has become an accessible and useful tool to predict trends. Just a search on Google data can reveal what users are looking for online, where the interest is growing and what they click on. For not being in a controlled environment, this data is as close as it can be to the actual consumer behaviour and it is easily quantified. Lyst is a successful platform that has been offering trend forecasts based almost exclusively in big data.

But big data shows what consumers already know something about, once they are already searching for it and the information is already somewhere online. In big data analysis, it’s important to don’t confuse trend spotting, which is what is on-trend now, with trend forecasting, something that is yet to come.

Then there is research, which can have many forms: a survey with consumers, a look on Instagram’s timeline or analysing the trends from other industries, like technology or entertainment. This tool has a scientific basis, but there is also subjectivity from the person conducting the research.

Finally, we have intuition, the feeling someone can have that a certain thing may happen. Even though intuition in trend forecasting is usually based on research and data analysis, it is a very subjective tool. Finally, with the use of statistics, projections, cause and effect comparison and other techniques that are often not clearly disclosed, trend forecasters manage to offer their predictions.

Trend forecasting has definitely developed more towards science in the last years. Marketing and consumer research became more sophisticated, especially due to big data and artificial intelligence. However, there is still an artsy subjective side to it. Louise Stuart Trainor, a consultant for WGSN, stated that trend predictions “used to be mostly art or, rather, intuition. But with the rise of big data, brands started to look for more hard evidence to bolster the concepts trend-forecasting agencies were providing.”

There are also those trends that come and go as cycles, such as the constant change between fashion minimalism and maximalism. For these, a study of the past can very often predict when they will repeat themselves. However, most trends are not as predictable just with an analysis of the past.

There is no question there is science in trend forecasting. At the same time, there is subjectivity on the creation of predictions, especially in the fashion industry where consumer decisions are very often based on abstract motivations. But these forecastings are rarely questioned in fashion. In the economics field, predictions of the future are looked with scepticism for years now.

40 years ago, the economist Friedrich Hayek made a bold speech during his acceptance of the Nobel Prize for economics. He stated that not only were economists uncertain about their forecasts, but their tendency to present their findings assuring the language of science was misleading.

While economic forecasting is based on scientific methods, a minimal change in a few variables can make the predictions too complex, close to impossible, to be done. Out of the 150 last recessions, economists failed to predict 148. The stock market predicted 9 of the last 5 financial crises. We are not good at forecasting and, according to Mark Pearson, deputy director for employment, labour and social affairs at the OECD Paris; we are getting worse at it. The complexity of the world is making even harder to predict the future. At the same time technology added accuracy to trend forecasting analysis, it also made the overall scenario more complicated. Predicting what people will want in the future is getting tougher and identifying what people want now gets irrelevant too fast.

An example of this complex scenario of constant change is high street retailers that depend a lot on trends. The most successful ones, such as Zara, are not the ones that predict trends before, but those that are highly responsive to them. The demand for newness is “less about forecasting and more about responsiveness.”

Finding patterns is not the hard part, the difficult stage is to identify amid these findings what is just noise and what is a signal (from the book The Signal and the Noise: Why Most Predictions Fail — but Some Don’t by Nate Silver). Thus, the use of intuition is indispensable in trend prediction this is also not exactly science. Trend forecasting is a form of art that uses science to achieve its aim — but then, in human science, what isn’t?

Trend Forecasters or Trend Makers?

Fashion trend forecasting has another phenomenon that reinforces the scepticism towards this industry: the self-fulfilling prophecy. To explain what it is, I will first present two situations outside the fashion industry.

First situation: If a meteorologist gives a weather forecast of rain, the fact that many people will carry an umbrella that day does not make it actually rain. The weather is not affected by how people reacted to the forecast.

Second situation: If a respected investor says the price of a stock will go down, many will decide to sell this stock to avoid losing money. Because people sold the stock based on the investor’s prediction, the stock value does end up going down. His forecast contributed directly to the prediction becoming true.

In the first situation, the forecast doesn’t influence on the actual event of the future. The accuracy of the prediction just becomes clear whenever the event happens. Differently, in the second scenario, the forecast influenced directly the event in the future. The prediction’s accuracy is irrelevant because it was the prediction itself that evoked a new behaviour that made the forecast (that could be false) become true. The second one is a self-fulfilling prophecy and it’s hard to argue this it is not happening in the fashion industry.

The trend forecaster WGSN is the platform used by almost all high street brands and fashion houses. It offers trends predictions for colours, fabrics, and design details up to two years ahead. In fashion companies, buyers and designers analyse WGSN’s content and, very often, base full collections and buying strategies on it. With so many companies looking at the same trend forecasts and then basing their decisions on it, it is not surprising when these projections become reality. WGSN is not predicting trends, it is making them. They work with self-fulfilling prophecies and by their slogan, which is “creating tomorrow”, they are probably aware of it.

And WGSN is not alone in offering self-fulfilling prophecies. Many other players in the fashion trend forecast industry do the same. Pantone, for example, releases annually the colour of the year, which often ends up becoming a trendy colour because brands trust Pantone and rush to have products with it. Even David Shah, editor and publisher of the biannual book of fabric palettes called Pantone View Colour Planner, admits this is happening, “Because of our platform, we’re able to promote (…) And it’s self-perpetuating — when people believe you’re right, they buy you. You get to be right!” he said. Some people even question how ethical this scenario is, but the truth is that it is the norm now. Mahir Can Işik, a successful buyer and merchandiser, brought some very good insights about this subject in his TED talk.

Trend forecasters are not alone in keeping self-fulfilling prophecies happening. As someone that worked in buying and design at a large fashion retailer, I can say all of us in the industry are guilty of maintaining this phenomenon. In a way, these platforms bring some safety to fashion companies and help designers and buyers to justify their choices. Even though they may be making the trend instead of forecasting it, they are still giving a small guarantee that the bets of fashion retailers and brands will be more assertive. At the end of the day, do brands really care if trend forecasters are making self-fulfilling prophecies if the other options would make the business riskier and harder to predict?

However, as highlighted by Mahir Can Işik, the fact that everyone follows the same predicted templates, which creates an environment where everyone has the same ideas. Thus, fewer people innovate in the industry. Fashion trend forecasting brings safety by predicting (creating) the future, but part of innovation comes from offering exactly what is not expected. This excessive use of fashion predictions is a dangerous path in which brands can lose differentiation — therefore losing competitiveness.

Marc Worth, the actual founder of WGSN, who sold the company before opening a competitor, often states he created a monster:

“Thousands of companies are signed up to WGSN, looking at the same colour forecasts, the same material swatches and the same silhouettes (…) It used to be a real source of inspiration to designers, but now it’s just doing their job for them (…) It has made life too easy for people in the creative space; it has made them lazy.”

Worth may be harsh with his words, but they have some truth to it. After selling WGSN, he opened a rival called Stylus. According to him, the platform offers market research and advice to businesses, but it doesn’t forecast anything. In his opinion, predictions became impossible to be done due to the changes that social media and technology brought to the market.

It’s Not About Predicting; It’s About Offering Possible Scenarios

As the mathematician, physician, and science philosopher Henri Poincaré once quoted: “It is far better to foresee even without certainty than not to foresee at all”. Yes, trend forecasting has many issues and it’s not as accurate as most may think it is, but by no means, I am saying it is not necessary. We can’t expect anyone to be able to precisely predict the future, but rather demand some likely future scenarios for companies to prepare for.

Marketing and consumer trend forecasting are a starting point for professionals to plan for the pessimistic, most likely, and optimistic scenarios of what’s next. It’s about thinking and preparing for the future, not perfectly predicting it. Getting predictions wrong is part of the forecasting business because no one has a crystal ball. There is no scientific method or form of art that can perfectly foresee what is yet to come. Maybe that is just not as clear as it should be sometimes.

With the Covid-19 impact on the fashion industry, the demand for trend forecasting platforms grew even more. Brands and retailers are desperate to know what they should do next, because the reality is that now, more than ever, no one knows the future of the fashion industry.

And for the forecasting of fabrics, colours and other garments’ characteristics…I hope it became clear those are not necessarily predictions. Brands that solely rely on these forecasting platforms are becoming boring for consumers. In the long-term, brands with a strong identity and innovative ideas are the ones that will stay successful.

To finalize, I will leave here phrase of Nate Silver that I believe fits a lot this discussion about trend forecasting:

“Distinguishing the signal from the noise requires both scientific knowledge and self-knowledge: the serenity to accept the things we cannot predict, the courage to predict the things we can, and the wisdom to know the difference.”

Check more content or subscribe to moderated newsletter by clicking here.

About the author behind the text

About the artist behind the illustrations

If you want, send suggestions, comments or a hi to luizapplima@gmail.com