
As fashion forecasting becomes even more imperative, predicting the future demand for particular products is also proving more challenging courtesy of rapidly evolving consumer tastes. In what can be a very volatile environment, how can companies get their supply chains in shape so they can best act on demand planning insights?
It is now financially and environmentally risky to over-order or overproduce. Aside from the sheer cost of production and shipping, companies also face the potential for markdowns if supply greatly exceeds demand. Exhibitors scheduled to present at the next ShopTalk tech conference discuss the ways in which brands and retailers can tackle the problem.
“The problem that we have these days is just moving things around the supply chain is becoming more expensive,” Iain Nicol, vice president of merchandising solutions at demand and supply chain software firm AGR Dynamics, said. “The more you order, the more money that you’ve got tied up in inventory; The more money you’ve got tied up in inventory, the more you’ve got to sell. And because of the fickleness of the buying public of the market, it’s very easy to end up with too much and therefore to end up with having to mark down.”
Fashion can be a tricky category to forecast, as the products are always evolving. Save for predictable basics, styles that are sold one season are usually replaced with new SKUs the next. Producers and retailers are therefore left with scant historical data on trendier garments and accessories. “You have less data, there’s less time to accumulate the data, and then the data doesn’t stay relevant for extremely long periods of time,” said Mike Shapaker, chief marketing officer at ChannelAdvisor, an e-commerce cloud platform.
As companies aim to avoid overstocks, there is also a risk of understocks. If there is a sudden surge in demand, retailers face the potential for missed sales opportunities if they cannot meet these supply needs in time.
For instance, through social media, trends can disseminate faster than ever before. A influencer post can cause an unexpected spike in interest for a particular style. If brands are partnering with influencers and providing outfits for them to model, AGR Dynamics can use artificial intelligence and machine learning to predict similar items that will rise in popularity based on the chosen look. Retailers can then prepare for the expected increased interest with healthier orders.
With certain timelier trends, speed to market differentiates whether a company is able to capitalize on excitement around a particular item or whether it will be too late. If a firm’s production lead time is four months, it may miss the season once it identifies a trend.
“Companies are competing on time,” Shapaker said. “It’s not even necessarily costs in all cases…And if you can win with the time aspect, it gives you a lot more flexibility to experiment and to really get your forecasting better.”
One way for companies to experiment and make up for the lack of existing sales data is by pushing products out onto marketplaces or e-commerce. Shapaker noted that this would enable companies to test out consumers’ real-world reactions to different colorways or style variations. Provided a brand has the agility, it could then ramp up production based on performance.
According to Nicol, while supply chains tend to be fairly sophisticated, what is often missing is “interoperability.” “Their internal processes are very slick, and they understand what they need to do, but the system that supports those processes can be very disjointed,” he said.
Rather than separate platforms, such as spreadsheets for some tasks and an enterprise resource planning (ERP) system for others, Nicol proposes a single system that can deliver timely, consistent information.This can also feed employees alerts about issues or changes such as a late delivery, required replenishment or a markdown. There is also a need for more flexibility in the supply chain to enable companies to move merchandise to the locations where there is demand.
While data is valuable, it is only one part of the equation. It is up to employees to actually take the data and interpret it, choosing which trends to act on and how. “I think both the technology and data as well as the human part are so equally important,” Elizabeth Shobert, vice president of marketing and digital strategy at data analytics platform StyleSage, said. “Sometimes maybe the people part of it can get a little bit overlooked when you talk about data.”
Aside from production, achieving speed to market can also be hindered by a company’s organizational structure. Fast-fashion frontrunners have been among the first to figure this out.
Shobert noted that part of what enables fashion firms like Zara to move quickly on runway trends or influencer outfits is a flat structure where employees are empowered to make decisions. “Yes, you have to invest in the right technology. Yes, you need to make sure people are trained and understand how to use the insights, but let them make those decisions,” she said. “That’s really where I think the power is and where a lot of organizations have not utilized people to their full extent.”