The Fashion Industry Finds a Crystal Ball in Data Analytics

Today, the world thrives on the ease of access to information and communication. This new paradigm has also kept the fashion industry on its toes, with companies trying to respond to dynamic trends while keeping consumers enthralled with new designs. Moreover, customers today are well informed about latest fashion trends and styles, designs, quality and prices.

As a result, fashion brands find it tough to keep customers engaged. Demanding and savvy customers, coupled with the vagaries of fashion, pose a tremendous challenge to the fashion industry.

In a highly competitive environment, business analytics and market intelligence are now must-have tools. How can a brand predict if paisley will be the next trend, or if orange will be the new pink? Analytics can help with everything from predicting the preferred mix of colors and sizes to pricing strategies that work. Retailers such as Marks & Spencer, J.C.Penney, Kohl's and Kroger increasingly have been adopting analytics to address operational inefficiencies and cost savings, and with remarkable results.

However, the fashion industry is perhaps one of the slowest adopters of analytics as a business tool, as shown in a recent study by SAP.

The fashion industry is unique in its manner of demanding freshness every season. Product development teams involved in coming up with new designs — designers, buyers and marketing executives — grapple with balancing customer expectations with the multitude of inspirations available.

In fact, designers go through the motions of deciding the next season's theme. These designs should both appeal to customers and align with the brand design philosophy that has evolved over time. Buyers base their decisions on commercial details while evaluating market trends highlighted by the marketing team and new inspirations drawn by designers. More often than not, line review meetings witness design decisions being based on collective judgment or intuition, rather than industry facts.

Jerome Fisher, founder and former chairman of Nine West Shoes once said, "Line up 10 shoes and 10 buyers and let them pick the hot shoe next season; you will get 10 different answers." This insight reflects the common malaise of perception driving preference. How does one evaluate or predict the success of a design before it reaches production?  Line decisions hence become very critical as choices can make or break the season.

Typically, customers purchase a style based on the complete package of the brand, design elements and price. Their transaction history can explain design preferences that can support new product decisions.  Historical data helps fashion businesses analyze best-selling styles, and customer preferences on colors, fabrics, finishes or silhouettes.

When analyzed over a period of time, this analysis can help companies predict trends and customer preferences. But this depends on the quality of market data, data accessibility and the right tools for extracting and analyzing relevant information from this data.

The key challenges faced by the industry are:
  • Fashion is dynamic and it's tough to predict trends.
  • There is no guarantee that a high fashion design will be successful.
  • Simply carrying over a best seller may offer no freshness to the line.
  • Lack of market data makes it almost impossible to judge the success of a new inspiration.
  • Attribute-level data collation requires tremendous effort and is not real time.
  • Designers have to rely on their own judgement in selecting the right inspiration
  • Category owners are unsure of the season performance as in net margins, inventory turns and markdowns
How can technology help?
With a multitude of systems in the IT landscape, fashion brands face the tough challenge of integrating all information to aid informed decision making. Existing processes are mostly manual and offer only limited data to buyers. Some companies with advanced reporting capabilities offer access to real-time data analysis. However, there is little insight into why customers preferred a particular design over the other. No analytical system has been able to give conclusive evidence on what draws customers to one style over another.

Robust PLM systems have been gaining popularity across fashion brands. These platforms can define design lifecycle-related parameters holding product attribute information to help monitor new designs while getting all development information on a single platform.

Since styles are basically a collection of attributes, PLM systems operating in tandem with transactional information from ERP systems intrinsically combine design related information with sales data to identify certain attributes that have contributed to a style's success.

Analytics through business intelligence (BI) tools has been regarded as one of the defining technologies for this decade. Representation of the PLM and ERP information on a BI tool through visual interfaces such as dashboards makes analysis easier and standardized across functions. It can extract analysis from historical data and cut through the clutter of style and attribute data for better decision-making.

These fact-based insights on a single BI platform can help teams to get their thoughts aligned in unravelling attributes that have succeeded or failed. These attributes can collectively form the basis of new designs from either the best sellers or the core carryover styles. New fashion inspirations, too, can be validated from the historical analysis if they were used earlier to check their appeal to customers.

How can companies benefit from fashion analytics?
Fashion companies need a solution that will let buyers know what worked and what did not. An intelligent tool will extract the best-selling attributes while pointing out the unpopular ones. This way, the assortment can include a little more of the best-selling range and weed out the non-performers, making design decisions more accurate.

If these best-selling attributes could have regional preferences, assortments could be tuned for addressing geographical customer requirements, thereby keeping customers satisfied and loyal. Brands, by way of this iterative assortment planning, can expect styles to find acceptance in the market, helping to push category sell-through percentage and increase inventory turns. This could then cascade into planning for lower markdowns, inventory levels and fit into any organizations profitability strategy.

Rajnish Kumar is global head of softlines and L N Balaji is president of North American operations for ITC Infotech.
This ad will auto-close in 10 seconds