HIVERY

HIVERY is a pioneer of next-generation retail assortment strategy simulation & optimization analytics. These proprietary machine learning and applied mathematics algorithms were co-developed and acquired from Australia’s national science agency - CSIRO’s Data61.

HIVERY was founded on the vision that Data Has A Better IdeaTM - and we’re working together with our clients to uncover its full potential.

The retail industry has undergone significant transformations over the past few decades. From the traditional brick-and-mortar stores to the advent of e-commerce and now the integration of artificial intelligence (AI) and advanced analytics, the retail landscape continues to evolve at an unprecedented pace.

In the early days, retail was a straightforward process. Consumers visited physical stores, selected products from the shelves, and completed their purchases at the checkout counters. Retailers used simple performance metrics such as total sales or rotation numbers to manage their stores. However, this approach had its limitations. Understanding customer behavior, preferences, and shopping patterns was challenging, which is crucial for optimizing store performance and enhancing customer satisfaction.

The advent of e-commerce brought about a significant shift in the retail industry. Online shopping platforms offer consumers the convenience of browsing and purchasing products from the comfort of their homes. Retailers had access to a wealth of data, from browsing history to purchase patterns, enabling them to personalize the shopping experience and improve their service offerings.

However, this data's sheer volume and complexity presented new challenges. Traditional data processing and analysis methods proved inadequate in extracting meaningful insights from this vast pool of information. This is where AI and advanced analytics came into the picture.

AI has revolutionized the way retailers process and analyze data. With the ability to learn from data, AI can identify patterns and trends, make predictions, and provide actionable insights. Advanced analytics, on the other hand, allows retailers to delve deeper into the data, uncovering granular insights that were previously inaccessible.

Today, AI and advanced analytics are optional extras and essential tools for any retailer looking to stay competitive. They offer the ability to leverage store-level data, providing a comprehensive analysis of customer behavior at the push of a button. This capability transforms various aspects of retail, from SKU rationalization and product innovation to assortment and category strategies.

As we delve deeper into this report, we will explore how AI and advanced analytics empower retailers and CPG suppliers to collaborate more effectively, make strategic decisions, and ensure they have the right product in the right place at the right time. The future of retail is here, and AI powers it.

In the retail industry, the mantra "location, location, location" has evolved into "data, data, data." AI and advanced analytics have ushered in a new era where store-level insights are valuable and indispensable for success. These insights, derived from granular data at the store level, provide a comprehensive understanding of customer behavior, preferences, and shopping patterns, enabling retailers and CPG suppliers to make informed, strategic decisions.

Store-level data goes beyond performance metrics such as total sales or rotation numbers. It delves into the intricate details of customer behavior, capturing the nuances of shopper preferences at individual stores. This granular level of data allows for a more precise and targeted approach to assortment planning, SKU rationalization, product innovation, and other category strategies.

The retail landscape is undergoing a seismic shift, with AI and advanced analytics at the helm. McKinsey's report1 underscores the transformative potential of these technologies, revealing that tasks once considered unsuitable for automation, such as “applying expertise” and “collaboration” between humans and systems, are now achievable. These two “tasks” collectively, the shift could lead to a staggering over 100 % productivity gain, effectively doubling current industry capabilities and speed to market value.

One of the critical areas where this transformation is evident is in the realm of store-level insights. A case study involving ten stores illustrates the power of AI in validating intuitive hunches and driving strategic decisions. The experiment revolved around the behavior of a specific shopper segment, the "domestic premium shopper" at a popular convenience retail store chain. The hypothesis was that these shoppers were primarily focused on seeking the lowest prices and were less concerned with whether the product was on a warm or cold shelf. Moreover, it was believed that premium shoppers preferred immediate consumption.

AI and data were employed to test these assumptions, leading to an experiment that assessed the feasibility and potential impact of shifting larger pack sizes of domestic premium brands from the cold to the warm planogram. The experiment revealed that by moving around 40 percent of the Points of Distribution (POD) of a brand, specifically the larger pack sizes, would generate a 13.7% increase in overall category revenue and an amazing 9.4% increase in revenue for our client's brand. This shift opened up a promising revenue opportunity and maintained a feasible plan for store layouts.

The experiment validated the initial hunch about the domestic premium shopper, leading to strategic decisions that respected the buyer's original decisions and maintained the integrity of the store layouts. This case study demonstrates the power of AI in turning intuitive hunches into actionable strategies, opening up new avenues for assortment plans, and overcoming space constraints. It's a testament to the potential of store-level insights in revolutionizing retail, challenging old assumptions, and presenting new perspectives for increasing category growth.

Case Study 1: Optimizing Assortment Space Contribution Index with AI

The Space Contribution Index (SCI) is a critical assortment planning and optimization metric. It measures the proportion of linear shelf space a product occupies relative to its sales performance. However, calculating the SCI can be complex, especially considering the vast amount of store-level data that needs to be analyzed.

AI and advanced data analytics techniques have been used to streamline the calculation of the SCI. By leveraging AI algorithms with store-level data, retailers can get a much more granular and accurate picture of their product assortment optimization. This leads to better decision-making and more effective shelf space allocation, regardless of the number of store clusters analyzed.

For example, a retailer looking to optimize its cereal category used AI-driven clustering analytics to identify patterns and trends in store-level data. The retailer had over 18,000 stores in the continental United States with just four clusters. After analyzing all four clusters, significant suboptimal indexing was observed, ranging from as low as 45% to over-indexing at 145%. By understanding this with AI's speed, the retailer could develop targeted recommendations for reallocating space based on specific parameters such as minimum sales thresholds or store door count.

Case Study 2: Tailoring Assortments to Individual Stores with AI

Understanding each store's unique needs and preferences is crucial for maximizing sales. AI offers the ability to prepare store-specific planograms based on store-level insights.

A recent analysis of 400 stores discovered that by using AI to create store-specific assortments, 8% of the stores (33 stores) could see an increase of over 15% in total incremental sales. However, as the retailer expanded store-specific assortments to more stores, the contribution to incremental sales began to decrease.

This analysis revealed a critical insight: while store-specific planograms can drive significant growth, it's crucial to select which stores to focus on strategically. In this case, the 33 stores with a potential sales increase of over 15% would be the best candidates for store-specific planograms. These stores likely have a unique shopper profile that can be effectively targeted with a tailored assortment.