How Machine Learning Transforms Assortment Optimization Approaches
In the ever-evolving world of retail and Consumer Packaged Goods (CPG), professionals like Retail Buyers, Category Buyers, Merchandise Buyers, Directors of Category Development, and Directors of Category Management face a daunting challenge: making strategic decisions about product assortment and space, considering a multitude of variables. The solution? Embrace the transformative power of machine learning. Let's delve into this groundbreaking approach.
The Essence of Assortment Optimization with Machine Learning
Assortment optimization using machine learning transcends traditional analytics. It leverages advanced algorithms to sift through and make sense of vast datasets. This technology enables retailers and CPG brands to curate the optimal product mix for each retail location, aligning with consumer preferences, local demographics, and purchasing behaviors. It's not just a change—it's a retail revolution.
Machine learning analytical models can leverage store-level data to gain insights into every product SKU and learn the optimal assortment combination needed while considering:
- Space Limitations: The Machine learning models allow users to explore assortment scenarios while considering space limitations at the store level.
- Shopper Preferences: It analyzes shopper data to provide hyper-local, retail-specific insights, enabling precise assortment plans per store
- Historic Sales Data: By leveraging historical store-level sales data, Machine learning models can offer dynamic assortment recommendations and SKU rationalization at the store level
- Fixtures and Merchandising Styles: There are real-world realities regarding the effective and efficient execution of optimized assortment plans. Machine learning models can evaluate store attributes such as fixtures and merchandising styles to provide dynamic assortment recommendations and maximize space productivity that is actionable to individual stores' needs.
What we have heard from customers
In our journey with numerous customers, we've identified recurring challenges:
1. Impact of Shopper Behavior Shifts and Supply Chain Issues:
The last few years have witnessed significant shifts in shoppers' buying behaviors and challenges in the supply chain. Retailers are concerned about adapting to these changes, including the rise and impact of omnichannel shopping and the challenges of managing inventory across different channels.
2. Data Overload and Making It Actionable:
CPG suppliers and retailers have access to vast customer and sales data but often need help to make this data actionable. The data often exists in silos and various forms, making it difficult to derive meaningful insights.
3. Integration of AI and Advanced Technologies:
There is growing interest in how machine learning algorithms and custom modeling can be utilized to manage the complexities of retail inventory and shopper data at the store level. Retailers are curious about how these technologies can be integrated into their existing systems without causing major disruptions. Machine learning analytical models used in HIVERY Curate can help. It uses store-level data to gain insights into every product SKU in every store to learn optimal assortment combinations and merchandise constraints, such as limiting assortment churn (it swaps to any percentage desired by the retailer or category management team to help limit operational disruptions).
For example, in customer analysis for a major breakfast and cereal manufacturer, our machine learning models identified just 10 SKUs that made up 20% of revenue while most efficiently ensuring days of supply (DOS). With this insight, the team discussed the possible assortment rationalization and tested scenarios around DOS thresholds, warehouse inventory, and distribution center (DC) constraints.
4. Aligning Metrics for Success:
Retailers face difficulties in aligning various metrics of success, such as weeks of supply (measured in time), inventory investment (measured in dollars), and service levels (measured in percentages). They seek ways to reconcile these metrics to achieve their business goals. Machine learning models allow users to set product assortment and store attribute constraints while considering business goals like revenue, volume profit, or a combination.
Change is constantly happening; thus, assortment optimization using machine learning is not just an option but a strategic imperative for retail and CPG brands. With machine learning analytics solutions like HIVERY Curate, Retail Category Buyers and Directors of Category Management are equipped with a powerful ally in their quest for retail excellence with speed and transparency that is actionable on the shelf.