(732) 400 5521

Data Assets for Fashion Trend Forecasters

By Catherine McClelland
Friday, October, 23 2015

Marsala—it’s the 2015 color of the year, forecasted to be the top shade of the Fall/Winter 2015 fashion season. The Pantone Color Institute singled it out in its semiannual Color Report, endorsed by all the big hitters: New York Fashion Week, Elle, Harpers Bazaar, Refinery 29 and Pop Sugar, to name a few. It’s already splashed across fashion magazine headlines and autumn look-books.

 

It’s a small company in northern NJ worth only $180 million. How do they set the color palettes for the most profitable players of the fashion industry?

 

Pantone’s methods and meetings are top-secret until the reports are released, but it’s clear that Pantone 18-1438 (Marsala) is no accident. The color industry is piloted by professional full-time colorists and trend forecasters who spend years designing and divining the shades that we will all end up wearing, from haute couture to bargain basement.

 

As Meryl Streep says in The Devil Wears Prada in the famous cerulean sweater scene, “That blue represents millions of dollars and countless jobs, and so it's comical how you think you've made a choice that exempts you from the fashion industry when, in fact, you're wearing the sweater that was selected for you by the people in this room.”

 

Forecasting is a highly manual process, and must be done with enough lead time to complete the entire design, production, marketing, and launch process. Trend forecasters have traditionally pulled from disparate sources to make their taste predictions. I recently talked to one forecaster who had attended a punk goth concert.

 

“I saw a pattern in the shirts and the face makeup of some of the audience members,” she said. “I knew I had something.” She based one of her major trend predictions on what she saw there.

 

Weeding through raw data is a process that takes place inside the forecaster’s intuition, but new data tools are on the horizon.

 

The key is that when it comes to getting an edge on the competition, the end product is less relevant than the customer. Companies that know their consumers on an individual level make more money than those that only track product data.

 

Customer data is difficult to accurately capture and leverage (think targeted ads and marketing initiatives based on predictive analytics), which is why it’s very profitable for those companies who successfully do it. Applying an analytical engine to solid data reduces the need for guesswork and specialized experience.

 

The theory of data analytics is that processes involving human intuition can be made more effective and accurate through the analysis of large amounts of data. It has everything to do with trend analytics and market power. Large fashion houses have the market share and resources to set trends, but the fastest growing companies are the ones who can use their data to more effectively tap trends.

 

The industry already understands the difference between straight-up data and a data asset. This term is often misapplied to raw data. I use it here to mean raw data that has been aggregated, quality-checked, sorted, and groomed into a set that is useful for drawing conclusions.

 

Broadly, the data assets for apparel retailers can be divided into product data and customer data. (We’re excluding manufacturing data to focus on apparel at the point-of-sale stage).

 

Product data assets are compiled from the raw data of sales figures, historical trends, and apparel attributes. Fed through a data analytics engine, they hold the key to regional trends, seasonal fluctuations, effectiveness of coupons, performance of certain colors and styles, and other metrics. Most retailers already track this at a very granular level.

 

The customer data asset is murkier, because until recently, the data simply wasn’t available. Companies began aggregating data such as gender, age, and buying patterns with the advent of credit cards and ecommerce. Now with the technology used by social media platforms, ad agencies and embedded marketing, it’s feasible to create a customer taste profile tailored accurately down to the individual level.  This raw data pool—albeit expensive and not always readily available—is now a reality.

 

Additionally, the fashion industry comes with challenge of regular “disturbances” to customer buying cycles: temperature changes, holidays, clearance pricing, and other influences on demand. Even TV shows (Mad Men, Downton Abbey, Peaky Blinders), have resurrected fashions previously thought to be done for good.

 

Turning customer data into a data asset is more difficult than just slicing, dicing, and organizing by metric. Products are simpler than people: fewer attributes, readily quantifiable, and no will of their own. Human tastes and decisions are more difficult to turn into data because of the number of unknowns that drive purchasing choices.

 

SAP has made strong inroads with its Demand Planning software. Its analytics engine loads actual or planned data from the company’s source and extrapolates based on statistical forecasting, advanced macro techniques, and take into account marketing intelligence and promotions. Other ERP software is following suit. Future modules will incorporate integration with granular and ever-evolving customer data at the individual level.

 

The apparel industry is sold on data. It’s too profitable to ignore. Big retail has already adopted modern data collection and analytical techniques for its product data, and now the competitive edge rests with the firms who can capitalize on the new frontier of available data: the customer data asset.

 

Trend forecasters can set themselves apart by asking their CDOs for the right data. Data teams have access to company sales and customer data, but they don’t have the specialized trend knowledge required to ask the right questions. With a stronger emphasis on forecasting via data query, forecasters can improve the accuracy of their decisions. Additionally, trend forecasters can gain the trust of non-design decision-makers when they can show data evidence for their decisions.

 

Trend forecasters aren’t going away, but their role is changing. They stand to gain strongly from the data revolution. Their key strength lies in understanding how data can benefit them, and using company data assets intelligently to augment their decisions.