As supply chains continue to grapple with volatility and disruption post-pandemic, fashion brands, who faced increased risks during the pandemic because of lockdown, are now turning to better supply chain planning for fashion strategies with the end goal of achieving improved agility and a competitive advantage. Brands like American Eagle Outfitters focused on inventory optimization and flexibility, supply chain transformation, lower SKU counts and fewer shipments per order for improved supply chain performance.
The pandemic accelerated digitization plans across multiple industries according to the ToolsGroup/CSCMP 2021 Digital Transformation in Supply Chain Planning survey. Supply chain leaders are leaning on technologies to keep up with evolving customer behaviors– the top driver of supply chain digital transformation. Armed with the right supply chain planning technology tailored to their industry needs, fashion brands can increase customer service levels, reduce inventory and unsold end-of-season items, reduce delivery time, and keep inventories lean while maximizing sales.
Synchronizing demand and supply across a dynamic global supply chain requires the ability to manage the disparate needs of multiple retail channels, seasonal demand, fashion changes and restrictive capacity constraints. Also, most fashion businesses combine rapidly changing seasonal collections with longer lasting continuous products, creating at least two different parallel demand streams and inventory processes, each with its own business logic. So we thought it would be good to examine some key capabilities in fashion supply chain planning and the technologies that support them.
- Seasonal merchandise planning for the collection – As firms selling fashion-oriented items plan their collections, demand planning software and analytics can help identify trends and potential demand from social influencers and the Internet. Since demand forecasting for fashion is tough to do based solely on “sensed” data without a sales history, firms can also identify merchandise from previous collections and sales patterns with similar attributes. Then, using web analytics and “weighting” these attributes, they can predict probable performance of the product launch, creating more detailed demand profiles for individual categories, subcategories and brands.
- Determining the assortment at the individual item, channel and store level – Fashion success depends upon splitting the demand forecasts into higher levels of assortment detail including material, size, color and destination sales channels. Demand modeling systems with attribute-based planning can disaggregate the merchandise plan down to the stock unit and item-location level, forecasting the expected demand that determines the product assortment presented in each type of store. For instance, stores located in malls might get one assortment; “high street” stores another. This is especially crucial for “pre-packs”—combos of the same items in different colors and sizes—so their initial allocation matches actual customer demand both for the items themselves, as well as what stock needs to be available in inventory—both in the store for on-site customers and in a distribution center for online shoppers. After fulfillment, inventory planning software refreshes the stock by continuously calculating optimal service points to satisfy the demand as it shifts between channels to minimize markdowns and obsolescence.
- Operational planning during the season for rapid replenishment of inventory based on sales signals – Rapid replenishment relies on a tight linkage between demand planning software and daily point-of-sale data to continuously break down the forecasts into channel and store-specific replenishment suggestions that capitalize on narrow selling windows. Machine learning in retail can help automatically adapt ongoing forecasts by performing demand sensing that drives timely in-season replenishment. This deep learning can “decode” the demand related data—structured and unstructured (social media, web)—analyzing the complex variables, interactions, and patterns to grasp the expected in-season demand profiles by item and by store.
- Margin optimization via timely inventory re-balancing between channels – Taking the pulse of sales also reveals opportunities for optimizing margin by avoiding stockouts and overstocks. Fashion includes products that don’t move quickly, but are important to profitability. Traditional demand planning will forecast safety buffers based on normal demand patterns. But this doesn’t work for slow-moving items. More advanced demand planning can model the probable “shape” of these slow movers’ demand, creating accurate demand and inventory models to manage stock for reliable service levels.
- Highly scalable modeling is crucial in multi-channel, where the software needs to process a large range of items that can proliferate extremely fast in a collection and down to the store level. This is true for both demand modeling (as described above) and for inventory optimization which builds the right mix of inventory needed to serve distribution networks that can scale to up to millions of SKU-Locations. It helps deliver high in-season customer service levels with limited late-season discounting that cuts into gross margins.
One European intimate clothing and hosiery apparel manufacturer is a great example of a company that faced multiple brands, products numbering in the tens of thousands, seasonal demand and many fashion changes. As a result, procuring materials from suppliers, factory production and warehouse logistics had become much more complex. They needed the right mix of inventory to serve many retail channels and increase customer service, and had to minimize unsold stock at the end of the season. Also, the company’s labor-intensive monthly planning process took up to 20 days. Using a demand modeling and inventory optimization system to speed planning and optimize margins, the company increased its customer service levels by 10 full points—to 96%—while reducing inventory by 15%. Read the full story here.
By Lorenzo Trucco