The complicated world of the merchants
The merchant role has become increasingly complex in the last decade. With increasing competitive and financial pressures, it has become imperative for retailers to differentiate themselves through personalized approaches and strategies, and to offer curated shopping experiences that are tailored to shoppers’ needs. To achieve this type of personalization, merchants must put data at the heart of their strategies. Grocery retailers are supreme examples having troves of data. Given the propensity for a customer to make multiple trips to the store and e-commerce sites for their daily or monthly needs, grocery merchants are privy to a lot of transactional data and digital footprints that customers leave behind. Right now, despite the abundance of data, retailers are only able to generate transactional reports. The challenge is to determine how to lure customers with compelling offers and new propositions like subscriptions, and to encourage buying across different grocery segments. There are added complexities too. What goes into the customer’s grocery basket is determined not just by individual preferences but by that of an entire household.
As Shelley Bransten, Corporate Vice President at Microsoft, says, “Having consumers in the driving seat is a big challenge. It requires understanding and moving at the right pace to meet the consumer of the future, and really getting to grips with where those consumers are going and what they expect. The next generation of consumers expect personalization at an unprecedented scale. Tastes and preferences are no longer declared but learned in real time through the ‘digital exhaust’ they create as they navigate their virtual and physical worlds. That’s going to put massive pressure on retailers and the technology they have in place today.”
As consumers increasingly adopt omnichannel shopping, merchants must weave in the context and shopper intent, as well as extend personalization to stores. This requires dynamic and integrated decision making across the value chain, factoring thousands of internal and external factors. Algorithmic retail is a paradigm shift in how retailers do business. By seamlessly integrating data across the retail value chain, they can unlock exponential value from retailers’ growing data assets by automating basic processes and adding intelligence to decision making processes across many functions and areas of the business, allowing key stakeholders to focus on top-level strategy. The real value of algorithmic retail can be derived when these capabilities coexist in harmony with humans, significantly amplifying and augmenting human ingenuity.
The opportunities offered by the digital merchant
What merchants lack is a comprehensive AI-powered, cloud-enabled platform that can help them analyze customers’ buying behavior and patterns that emerge from repeat purchases across channels and drive better customer-centric strategies.
The digital merchant is a concept that many retailers are pursuing these days. They want a cockpit view of their operations that is data driven to help merchants drive real-time decisions across the value chain. For example, if a retailer decides to run a long-term promotion on the cheese category, what is going to be the implication on the facings required for each item to support the increased unit movement? How do these increased facings roll up to the total space required for cheese? How will it impact the performance of the overall dairy category? Will it be necessary to pull space away from other categories and make room? When the promotion finally ends, what should the regular price be?
To effectively actualize this concept, we need to deliver the promise of the digital feedback loop. It is the capability to connect data from different parts of the organization for cost savings or from a top-line perspective across each business area. In the cheese example, a retailer would need a strong foundation in price optimization, competitive pricing insights, assortment optimization, and potentially space optimization in order to deliver.
Price, promotion, and markdown optimization
Pricing has long been determined by a series of rule sets, often managed in Excel. These rules rely on very large sets of data including customer data, performance/transactional data, store location/zone, and elasticities. Therefore, harnessing these successfully to optimize pricing is beyond the capability of spreadsheets. True price optimization allows retailers to balance all the complexities of price setting and maintain control over margins over their pricing strategy, and if done well, their customers’ price perception.
Related areas to price optimization are the promotional and markdown optimization functions. Many retailers still use manual tools and processes to track these tasks. One of the problems at the heart of promotional optimization and markdown optimization is that many retailers revert to averages when analyzing and planning the next event, resulting in massively suboptimal results. They also don’t have easy-to-use, user-driven forecasting support. In both cases, there could be elements of vendor deals and vendor funding that may influence the best packages and discounts to offer. There is also the view of how much inventory is left across the stores and distribution centers and when the product should be exited or how long the promotion should run. So, with the complexity of performance, vendor contributions, time, and anticipated customer response, price optimization is a key focus area for retailers wanting to leverage artificial intelligence (AI) and machine learning (ML) to model and recommend strategies.
Space optimization happens at many levels in most organizations. Typically, the first level is an optimization run that will help to decide fixture plans (freezers or gondolas). This run is often executed on square footage space to allocate the top-down view of a floor plan and is most applicable for new stores or remodeled stores. The next step is typically to optimize space within the department, letting the categories compete for linear space. Here, retailers are looking at how they should allocate space for maximizing shopper impact within the stores.
Assortment optimization is arguably the most data-rich decision that a merchant must make. This decision combines all the datasets previously mentioned and will dictate how the total strategy reaches the customers. To make the best decision, a merchant needs a mechanism to weigh all the inputs to determine the most profitable, productive mix for each store or group of stores. Additionally, the merchant needs to apply visual and strategic constraints onto the optimization to help the algorithm produce an assortment that is shoppable and understandable by the customers. This can be a rule to ensure that the assortment includes items from the good-better-best brands, or a rule that there can only be large bags on the bottom shelves due to fixture weight capacity.
The merchant’s personalized view
While there are productivity and tool advancements in the areas of pricing, competitive pricing, assortment optimization, and even space optimization, these activities remain largely isolated within the reset calendar. If we go back to our cheese promotions example, we can see that if the questions raised are answered in isolation, they will only give a point-in-time answer. But, if looked at concurrently, a system can help recommend a more impactful and profitable strategy.
Personalization is not just for consumers; merchants want it too. In addition to the power of concurrent optimization, the digital merchant can also solve many of the process challenges arising from different business types and different styles of the merchants as individuals. For example, a meat buyer needs to analyze the category with variable weights, whereas a cereal buyer has a more consistent unit of measure for analysis. By putting all the capabilities in one place, a structured workflow can be established that not only satisfies the need to follow the reset calendar, but also ensures that there is consistency in the robustness of data and capabilities leveraged. Alongside the merchandising activities, having an intelligent workbench for the merchant also allows for better decisions and integrations with supply chain. Supply chain parameters such as replenishment frequency, picking multiples, and case packs can now be modeled and analyzed earlier in the process. The result is a more agile and nimble handshake between merchandising and the supply chain.
Playing intrapreneur and storyteller
Today, merchants are not only intrapreneurs responsible for the growth and profitability of categories, but also aspiring storytellers. AI and digital technologies can simplify a merchant’s life by helping them understand the preferences of each customer/household, which typically varies based on location, occasion, and environment. Additionally, AI learns and adjusts itself when customers start behaving differently. AI solutions that are not black box solutions help merchants to see the not so obvious relationships and derive interesting insights, helping them understand the rationale for the assortment mix and price points. This makes it easier for them to tell their category story of why and how items will drive traffic, complete baskets, and increase basket size for targeted customer segments.