Winning the Food Fight:Best Practices for Managing Grocery Retail Supply Chains

1. Retail Strategies Fall Flat If Not Backed by the Right Supply Chains

Food retail is a tough and turbulent market. Grocery has never been easy, but the current business transformation is more dramatic than anything we have seen in decades.

Grocery retailers need to simultaneously address several major trends:

  • Retailers face great pressure to stay abreast of developments in online ordering and the increasing range of order fulfillment options, from express delivery to curbside pickup. Online shopping is especially challenging for food retailers due to the low value of grocery products combined with the high cost of handling products that are often frail, come in various shapes and sizes, and require temperature-controlled storage and transportation. Currently, few players are able to make money on e-grocery, but no food retailer can afford to ignore the emergence of online.
  • Traditional full-assortment grocery retailers are feeling the squeeze both from discounters and the food service industry. Restaurants and food-to-go are capturing an increasing share of consumers’ wallets. Some food retailers have opened in-store restaurants. Even more are turning to prepared meals to increase their relevance in the eyes of the consumers. High-value food-to-go items also offer grocery retailers a means of increasing margins in a tough market — if done right. Done wrong, it also offers plenty of opportunities to lose money on costly food waste.
  • Consumers are increasingly looking for fresh and healthy eating, putting produce, fresh fish and meat and other short shelf life products in the spotlight. Even discounters, such as Aldi and Lidl, are boosting their offering of fresh products, including premium products like organic meat and freshly baked bread. These former hard discounters are, in their pursuit for growth, stepping out of their comfort zones by introducing more fresh products, smaller stores and more regional variation in their offering. Discounters’ efficiency had previously built on simplicity, standardization and large volumes, so this increasing complexity will put their grocery supply chains to the test.
  • Urbanization and convenience are pushing many former big box retailers to experiment with smaller store formats in urban locations. Other retailers are reallocating space in stores to make room for cafés and deli areas or to accommodate online order pickup. To make the tighter space work without slashing assortments, retailers need to become more efficient in their use of space. It is not a given that successful big format retailers will accomplish the mindset change required to run small format stores, especially to do it in a way that creates synergies between the formats.

Looking at these trends and the challenges and opportunities they present, it is obvious that supply chain management will lie at the heart of the future successes and failures in grocery retail. All food retailers need to make tough choices today about where to place their business bets. However, regardless of the strategies selected by the different players, if their grocery supply chains are not developed to match the chosen strategies, chances of success are slim.

Successful food retailers need to master both the hard discounters’ lean, highly efficient grocery supply chains as well as the agile, responsive grocery supply chains needed for fresh products. In addition, many of them will need to manage the complexity of operating multiple store formats and offering several fulfillment options in parallel.

Figure 1: Responsive vs Efficient

Figure 1: Fresh products with a high risk of waste require a more responsive grocery supply chain. For center store, ambient products and other products with longer shelf-lives, operational efficiency is key.

To achieve this, retailers need the right planning tools at their disposal. Furthermore, they need to understand how to apply them:

  • Fresh products typically have a high risk of markdowns and waste, making it very important to accurately forecast demand and replenish in sync with demand. The planning process for fresh products must be granular enough to capture even the smallest changes in demand. The supply chain for fresh products must be agile enough to adapt to variable demand.
  • Center store and other longer shelf life products are key to attaining efficient goods handling and optimized inventory flows. Accurate forecasting is essential, but supply does not need to be exactly in sync with demand at each individual moment. This makes it possible to level out the inventory flow through the supply chain for efficient utilization of capacity. In the stores, setting up supply to allow for one-touch movements or “truck to shelf” delivery is essential for increased profitability.

In this best practice guide, we will highlight key approaches for increasing both responsiveness and efficiency in grocery supply chains. You will be hard pressed to find a single retailer employing all of these best practices. Rather, we encourage you to prioritize the most feasible and impactful development areas from your own perspective.

2. Harness the Power of AI in Demand Forecasting

Demand forecasting is the engine running your supply chain. High quality forecasting requires making the most of all available data. The richer the data, the more accurate forecasts you can produce.

It is impossible to talk about demand forecasting without discussing artificial intelligence (AI). Recent breakthroughs in AI, such as a computer program beating a professional human player in the intricate game of Go – have given rise to a lot of excitement, but also a lot of hype and unfounded claims.

Within the supply chain arena, the boldest vendors’ marketing departments are presenting claims that AI is approaching singularity and supply chains are growing autonomous. In the context of AI, singularity means an AI of such intelligence and power that it starts to independently develop and improve to an extent that renders us inferior humans redundant in some apocalyptic future.

Claims of singularity and autonomy can safely be filed under nonsense. We are still far from some kind of autonomous AI even within the very narrow context of supply chain management. We are even further from cracking the code of general artificial intelligence.

What we are seeing is great progress in specialized AI. Specialized AI means methods and algorithms optimized to perform a specific task. The original AlphaGo program that managed to beat the best human players in the game of Go had been specifically optimized for playing Go. It was trained using a database of around 30 million moves. After this, additional data was accumulated by having several instances of AlphaGo play against each other in a set-up known as “reinforcement learning.” The latest version of the program, AlphaGo Zero, uses no human training data at all. Yet, the data collected by playing is still processed by man-made optimization algorithms specifically designed for great performance when intelligent search through an enormous space of possibilities is needed. AlphaGo’s techniques have successfully been applied to chess and other games but are less useful in domains that are difficult to simulate, such as driving a car. It is also important to keep in mind that the data processing required comes at a cost. The hardware cost for a single AlphaGo Zero system was estimated to be around $25 million in 2017.

robot and washing machine

Figure 2: Which is the better robot for washing clothes? (Illustration inspired by an excellent blog post by Ben Evans.) Specialized AI is growing increasingly common and is often used to run applications that at first sight do not look particularly intelligent.

Two factors are key to the recent developments in specialized AI: 1) The rapidly increasing availability of data and 2) the rapidly decreasing cost of processing data. The current boom in AI was to a large extent fueled by inspiring advances in computer vision. These advances, on the other hand, were enabled by vast amounts of classified images becoming available to researchers at the same time that the cost of, for example, random access memory (RAM) has plummeted.

Similarly, access to new data and affordable data processing has made AI technology a must-have in any retailer’s planning toolbox. Essentially, AI adds new, more sophisticated tools to your toolbox. These tools, such as machine learning algorithms, make it significantly easier to analyze very large amounts of data to identify new, sometimes surprising patterns or to detect patterns on a more granular level than ever before. Machine learning , for example, enables automatic estimation of the impact of weather changes, such as expected precipitation or temperature, when forecasting demand for specific products in a specific store on a specific weekday in a specific season (see Section 2.4 for more details).

Machine learning in demand forecasting is something you need to adopt in order to be competitive in food retail. However, you also need to understand its limitations. Automating the bulk of demand forecasting is both desirable as well as quite feasible in food retail. Yet, the business environment is very dynamic due to changing consumer trends as well as the impact of external factors, such as the highly unusual weather in several parts of the world lately. There is always a risk of forecasts being based on how things used to be instead of how they are now or will be in the future. For this reason, there will always be errors in the forecasts produced. For experts to be able to understand the errors, potentially correct them or at least predict when they will happen, transparency into how the demand forecast was formed is essential.

Although a “black box” forecasting system, taking in all sorts of data to produce the most accurate forecast possible, is an appealing thought, it may well kill your business or at least your planning efficiency. This is due to two reasons: 1) Occasional extreme forecast errors can be very detrimental to a retailer’s performance, much more so than frequent smaller errors, and 2) forecast errors that do not seem to make sense erode the demand planners’ confidence in the forecast calculations, leading to increased double-checking and manual forecasting and crumbling the whole purpose of harnessing computer power for forecasting.

Best practice retailers use all tools in the demand forecasting toolbox wisely, selecting methods not based on trendiness but accuracy and robustness. They are also constantly reviewing forecasting performance and errors to support further improvements.

No black boxes

2.1. Choose the Right Tool for the (Forecasting) Job

Different forecasting approaches have different strengths and weaknesses. Some forecasting methods may be highly accurate when given access to tons of data only to fail miserably when there is too little training data available. Others may be computationally very effective and produce results that are roughly right but never stellar. Some forecasting methods are invaluable for short-term forecasting but do not add value when focus is on the longer term. There is no such thing as one single best forecasting approach. In fact, it is often surprisingly difficult to even agree on one single best forecasting result.

The best practice in demand forecasting is to use a combination of methods, ranging from traditional time-series forecasting to machine learning. In this way, you can attain highly reliable and accurate forecasts at the same time and produce forecasts that work both for short and long-term planning.

When combining several forecasting methods, we recommend using a layered approach (see Figure 3). This means that different parts of the forecast, such as baseline sales and impact of weather, can be viewed separately. The layered approach creates transparency into how the final forecast has been derived, which in turn promotes understanding and confidence in the demand planners. It also supports error correction and continuous development of the forecasting methods in use.

Figure 3: Forecasting Layers

Figure 3: Typical forecasting layers. The three layers presented here are not exclusive; additional layers can be added when new relevant data feeds become available.

2.2. Baseline Sales

Time-series forecasting is a solid and well-understood approach for estimating baseline sales. By using a set of best practice statistical tests and time-series models, different kinds of sales patterns, such as trends, seasonality and weekday-related variation in demand, can be modeled accurately.

Time-series forecasting has been around for a long time, giving many people the impression that time-series forecasting is “old hat” and that its potential has already been exhausted. This is far from true.

Despite the current focus on sophisticated forecasting approaches such as machine learning, there are still a great number of notable grocery retailers who have yet to truly embark on their journey of data-driven forecasting. In a 2018 survey of North American grocery retailers, less than half of the respondents declared that they were able to produce forecasts on the day-SKU-store level (SKU = stock-keeping unit). The other retailers would have wanted to do day-level forecasting, but simply could not do it.

There are, thus, big differences in how well retailers have managed to implement time-series forecasting and, consequently, in the forecasting performance attained.

The best forecasting systems automatically select the optimal forecasting models and parameters per store and SKU. This is typically done based on an array of statistical tests which identify demand patterns, such as seasonality or trends.

Automated selection

Figure 4: Automated selection of forecast models builds on efficient data cleansing to remove non-recurring exceptions from the sales data, statistical analysis as well as optimization for minimized forecast errors.

In retail, a typical challenge in demand forecasting is low sales volume at the day-SKU-store level. It is of central importance that the planning system is able to automatically move between day-SKU-store and more aggregate levels as needed to ensure that forecasts are based on sufficient data.

Below are some examples of combining day-SKU-store level forecasting with forecasting on more aggregate levels:

  • Slow-moving products with seasonal demand patterns. For slow-movers, sales volumes on the day-SKU-store level may be too low to enable distinguishing seasonal patterns from random variation. Yet, when looking at the total sales of these products in hundreds of stores, the seasonal pattern becomes clearly visible. Moreover, the seasonal variation may be strong enough to warrant proactive stock adjustments at the distribution center level. Best-in-class planning systems can detect product-level patterns and apply them to the SKU-store level forecasts to ensure that the same forecast serves both store replenishment and purchasing to the distribution center.
  • Weekday-related variation in sales. In general, sales forecasting in a retail setting is best done on the week level to attain sufficient sales data to work with. However, especially for products with short shelf-lives, it is crucial to get the day-level forecasts right. Best-in-class planning systems use intelligent profiling to break down week level sales into day level sales. This way of forecasting also makes it possible to copy past profiles for specific weeks, such as the weeks around Easter for which the previous year is a good benchmark with the holiday always occurring on the same weekdays, or customize profiles for weeks where a specific date determines the start of a holiday, such as Christmas.
  • Intraday forecasting. For products that are replenished several times per day as well as for workforce planning, forecasts on the hourly or even more granular levels are needed. This can be attained by applying intraday profiles to the different weekdays.
  • Online orders picked in stores. When retailers pick online orders in stores, best practice is to generate separate forecasts for in-store and online sales. Although store replenishment is driven by total sales, resource planning for order picking requires its own forecast as there is usually differences in how in-store and online sales are distributed over the week.
    Data cleansing

2.2.1. New Products and Stores

As time-series forecasting relies on finding patterns in historical sales data, additional routines are needed for dealing with new products.

The most common approach is assigning a reference product to the new product to be used as blueprint for the new product’s sales pattern until it has accumulated sufficient historical data of its own. However, in sectors such as grocery retail, the number of new products per year can be massive. This means that manual identification and setting of reference products is infeasible or at least highly inefficient.

A much more efficient approach is to automatically assign reference products based on product attributes. Relevant attributes are, for example, product group, brand, pack size, color or price point. The same approach can, of course, be applied to finding suitable reference stores for new stores.

New product references

Figure 5: For a new product in the cereal category, a good initial reference product can be found by looking for products with the same brand, size, and differentiating features, such as organic and whole-grain.

2.3. Impact of Promotions and Other Commercial Decisions

An important source of variation in demand is a retailer’s own commercial decisions, such as promotions, price changes or changes in how products are displayed in the stores. Despite these changes being controlled by the retailer, their impact is in many cases not very accurately predicted.

In a 2018 study of North American grocery retailers, almost 70 % of the respondents indicated that they cannot consider all relevant aspects of a promotion, such as price, promotion type or in-store display, when forecasting promotional uplifts. But they wish they could.

The best practice for estimating the impact of commercial decisions such as promotions is to use multi-variate regression analysis. Multi-variate regression analysis considers any number of factors that may have an impact on demand and automatically identifies relationships between the variables and sales based on patterns in past sales.

Variables that can be considered include, but are not limited to:

  • Type of promotion, such as price reduction or multi-buy
  • Marketing activities, such as newspaper ads or only in-store signage
  • Price change compared to the product’s normal sales price
  • Display in store, such as presenting the promoted product on an endcap or a table

Endcap promotions

Figure 6: For this product, an endcap display with no price change results in a notable sales uplift, but the uplift is modest compared to the effect of the 50 % price reductions.

Multi-variate regression analysis also produces valuable information that can be used to create better performing promotions. The analysis increases your understanding of the price elasticity within your product range as well as your understanding of which promotion types, displays and advertising work best for your different product categories.

2.3.1. Cannibalization and Halo Effects

It is quite typical that a promotional uplift for one product results in reduced sales of another product. If a supermarket carries two brands of lean organic ground beef — HappyCow and GreenBeef — it is reasonable to expect that promoting the HappyCow product will result in more people buying it, but also in some of the baseline demand from GreenBeef shifting to HappyCow. If the demand forecast for the GreenBeef product is not lowered, there is a high risk of stock-piling leading to waste.

Cannibalisation

For most center store products, such as canned food or cereal, cannibalization is not a big problem. If demand decreases temporarily, a replenishment order for the cannibalized product will simply be triggered later than usually. However, when working with fresh products and especially products that have a limited number of direct substitutes, the cannibalization impact needs to be considered to avoid excess stock and spoilage.

Manually adjusting the forecasts for all potentially cannibalized products is infeasible in food retail due to the large number of products and typically quite store-specific shopping patterns.

Best-in class planning systems automatically identify cannibalization and adjust forecasts accordingly. This can be achieved using regression analysis to identify relationships between the sales of different products. If an increase in sales is correlated with a decrease in the sales of another product, the products are considered to be cannibalizing each other.

If, on the other hand, an increase in sales is correlated with an increase in sales for another product, the system has identified what is commonly called a halo effect. The halo effect can, for example, mean that a promotion for Bolognese sauce will have an impact on the sales of spaghetti.

2.4. The Impact of Weather and Other External Factors

External factors such as the weather, local concerts and games, and competitor price changes can have a very significant impact on demand.

It is often intuitively easy to understand how, for example, weather impacts sales. High temperatures increase ice cream sales, rainfall increases the demand for umbrellas and so on. However, when looking at the entire product range a retailer offers, it becomes more complicated.

How can you effectively identify all products that react to the weather? How to account for “weather” encompassing a whole bunch of different variables, such as temperature, sunshine and rainfall? How can you consider some weather effects being stronger in summer than in winter or stronger during the weekends than on workdays?

Quick reactions to demand

For a mid-sized retailer with 400 stores and a range of 10,000 products, considering weather effects on a reasonably granular level would mean examining the strength of 2.2 billion potential relationships between variables (400 stores x 10,000 products x 20 weather variables x 7 weekdays x 4 seasons).

The use of weather data and forecasts is a great example of the power of machine learning. Machine learning algorithms can automatically detect relationships between local weather variables and sales of individual products in individual stores.

In addition to mapping these relationships on a more granular and local level than any human would be able to do, these algorithms are able to detect less obvious relationships between weather and sales. In a manual process where demand planners or store personnel check weather forecasts and make decisions accordingly, focus necessarily has to be on securing supply when demand is expected to increase — for example by pushing additional ice cream into stores in expectation of a heat wave. Usually, though, no one has time to adjust forecasts slightly downwards when rainy and cold summer weather reduces the appeal of barbecuing.

In our experience, considering weather effects in demand forecasting reduces forecast errors by between 5 and 15 % on the product level for weather-sensitive products and by up to 40 % on the product group and store level.

As discussed in the introduction, we strongly recommend a layered forecast approach, which delivers transparency into the different components of the forecast. This is particularly important when using external forecasts such as weather forecasts, which include an element of uncertainty.

One way of achieving this is to offer demand planners two versions of the forecast: One demand forecast without the weather-related effects and another weather-based demand forecast incorporating the impact of the expected weather (see Figure 7). In this way, planners can decide on a case-by-case basis how much emphasis they want to put on the weather-adjusted demand forecasts in anticipation of, for example, a heatwave that might hit a region during the weekend.

Weather based demand forecasts

Figure 7: For this product, baseline sales follow a regular weekly pattern. However, based on historical sales patterns, the product has been identified to react notably to sunny weather on the weekends.

In a similar way, machine learning algorithms can be used to take advantage of other external data sources in addition to weather to independently look for relationships between external variables such as local football games and local sales of specific products.

In grocery retail, the following external data sources have been found particularly useful:

  • Local weather data and weather forecasts
  • Passenger number data and passenger forecasts for, for example, airports or train stations
  • Information on past and future local events, such as football games or concerts
  • Data on competitor prices

3. Refine Your Store Replenishment for Increased Availability, Reduced Waste and Maximum Efficiency

The quality of a retailer’s store replenishment process has a direct impact on its top line revenue and its bottom line profits.

A better store replenishment translates into the following benefits:

  • Increased revenue from better on-shelf availability boosting sales up to 1%
  • Up to 30% lower markdown and spoilage costs as supply matches demand more accurately
  • Optimized inventory flows enabling up to 30% reduced cost of goods handling in the distribution centers and stores
  • Much more efficient capacity utilization in transportation, storage and manual work phases throughout the supply chain

Yet, in a 2018 survey of North American grocery retailers, 20% of the respondents had not even started implementing any kind of replenishment automation in their stores and only 30% of the companies had implemented forecast-based automated store replenishment extensively. Store replenishment is definitely an area where many food retailers’ operations are quite far from best practice.

3.1. Fresh Food Replenishment Warrants Granular Planning and Control

For fresh products, well-managed store replenishment is central to finding the optimal balance between the risk of lost sales margins caused by stock-outs and the risk of waste or markdowns eating already slim margins.

Even though traditional supermarkets have decades of experience dealing with fresh products, many still do not excel in this area. Their supply chains are reactive enough to support frequent deliveries, but their replenishment planning is not up to scratch.

According to the North American grocers surveyed, the annual value of spoilage was on average around 70 million and up to several hundred million annually for the largest companies offering a wide range of fresh products. The survey responses also demonstrated that grocery retailers with less advanced store replenishment practices are almost five times more likely to suffer from above-average spoilage levels than their competitors who are using forecast-driven store replenishment extensively.

3.1.1. Balancing Waste and Lost Sales

For so called ultra-fresh products, meaning short shelf life items that need to be sold that same day, a 100% on-shelf availability means that there will always be waste or markdowns unless the forecast is consistently flawless on the day, store and product level. This means that very granular control is needed to find the optimal balance between the risk of stock-outs and the risk of waste. Other fresh products face a similar challenge, just a bit less pronounced.

Demand for a product in a specific store typically varies between different weekdays. For some stores and products, this weekday variation in fresh replenishment can be very dramatic. This means that the same safety stock does not fit all weekdays when dealing with short shelf life products.

Sales and baseline forecasts

Figure 8: In food retail, it is quite typical that demand for the individual items varies within the week. If the safety stock levels are static throughout the week, there is a high risk of stock-outs when demand is high or a high risk of waste when demand is low.

Roast beef, for example, tends to sell a lot more towards the weekend than after the weekend. For roast beef, a static safety stock level leads to 1) Excess inventory after the weekend, with increased risk of waste, and 2) perilously low safety stocks during the weekend, with an increased risk of stock outs.

To find the right balance between the risk for waste and the risk for stock-outs, safety stocks need to move up and down in step with the expected sales volumes and forecast errors for the different weekdays. Good retail planning systems do this kind of granular safety stock optimization automatically.

In fact, the best retail planning systems take optimization even further by not only enabling dynamic safety stocks, but optimizing each order based on cost-benefit calculations that balance the risk of waste against the risk of out-of-stocks. Such machine learning algorithms minimize the total of lost sales margin and cost of waste.

The cost function needs to be adjustable in terms of how much weight it places on on-shelf availability vs. waste to allow for considering the strategic roles of key categories and items as well as whether there are many or limited opportunities for substitution within the product category.

Total costs vs order quantity

Figure 9: Best-in-class retail planning systems optimize each order based on cost-benefit calculations that balance the risk of waste against the risk of out-of-stocks.

When managing store replenishment of fresh products, it is very important that all calculations and optimizations are done automatically. It is an impossible task for any human to keep track of all factors influencing demand, such as weekday variation (e.g. seasons, weather, and promotions) as well as all factors influencing replenishment (e.g. delivery schedules, batch sizes and day-level probabilities of waste and stock-outs) for hundreds or thousands of items per day in a store, let alone hundreds of stores.

However, it is equally important that the forecasting and replenishment system does not turn into a black box. Actionable analytics allow supply planners to easily detect and remedy exceptions such as historical or projected waste or poor availability.

Examples of typical exceptions in fresh food replenishment are:

  • Too large order batches causing waste in stores. Sometimes order batches, such as case packs, are so large in relation to a store’s demand that each delivery of the product will result in waste. To address this problem effectively, your supply planners need to be able to determine whether this is a problem in just a few or many stores, what the financial implications are, and whether the problem can be mitigated by directing replenishment to specific weekdays, such as ordering the products only for the weekend.
  • Too much allocated shelf space causing waste in the stores. Sometimes visual minimums, designed to keep displays attractively stocked, drive excess stock and waste of fresh products. Supply planners must be able to identify whether the problem is isolated to a few low-demand stores or whether there is a more widespread problem with the planograms in use.
  • Systematically poor availability or high waste on specific weekdays. Systematic patterns of poor performance on certain weekdays, such as higher than average waste on Mondays, is not uncommon. To address this problem, your supply planners need to understand the root cause of this problem. There can, for example, be process issues, such as store personnel checking sell-by dates and recording waste on specific weekdays, which need to be accounted for in replenishment planning.

Automation radically reduces the time spent on routine tasks in store replenishment planning. At the same time, it multiplies the impact of your most knowledgeable process experts. If store replenishment has not been automated, your best supply chain analysts have limited leverage. They can review successes and failures in the rearview mirror and try to turn a few of their findings into action in the stores with help of the field training team.

When store replenishment is automated and replenishment planning centralized to a knowledgeable team, your planning experts can make a visible difference in hundreds of stores, almost immediately, simply by fine-tuning replenishment settings.

3.1.2. Stores Turning into Kitchens

With consumers increasingly looking for convenience, food-to-go and meal solutions are on the rise. Many stores are turning into kitchens, where sandwiches, hotdogs, and salads are made.

Traditionally, products manufactured on-site have been considered special items that have to be managed manually in the stores. However, with the growing demand for ready meals, the importance of on-site production has grown much more pronounced and more critical to food retailers’ profitability.

The replenishment process of ready meals is not all that different from replenishing other products sold in a store. It is just a bit more complicated. The demand for the end products — the meals — needs to be translated into the ingredients used to manufacture the end products. The replenishment calculations need to be done for each ingredient accounting for each ingredient’s lead time and on-hand stock.

Products produced on site

Basically, the process is the following:

  1. 1. Forecast end product demand.
  2. 2. Translate estimated demand for end products into estimated demand for ingredients needed to manufacture the end product. This requires knowing the recipe (sometimes also called the bill-of-materials, a term borrowed from the manufacturing industry) as well as the yields of the different ingredients. If 1.3 oz of lettuce is needed for a sandwich, the calculations may need to be done according to 1.7 oz of lettuce per sandwich to account for part of the lettuce not suitable for use in a sandwich.
  3. 3. Calculate the estimated demand for each ingredient. The total demand for an ingredient often reflects its use in several end products.
  4. 4. Calculate the required replenishment quantity of each ingredient based on lead time, available stock, potential incoming orders, estimated demand and target safety stock.

Sometimes, the ingredients included in a recipe are composed of other ingredients, such as special mayo or mustard produced on-site. In those cases, similar calculations need to be conducted for several levels of recipes. A horrible task for any human being, but quite manageable for a computer.

Product bill of materials

Figure 10: Replenishment planning of meals produced on-site, such as sandwiches and salads, requires factoring in the recipe or so-called bill-of-material of each end product.

3.1.3. High-frequency Replenishment

For ultra-fresh products, many retailers have chosen to deliver them to stores multiple times per day to guarantee freshness. Similarly, items produced on-site are typically prepared in several batches during the day. This applies especially to the growing category of in-store bakery products, which ideally should still be warm when the customer picks them up. In addition, the new trend of food retailers opening small stores in urban locations has made several replenishments per day a must due to lack of in-store storage space.

Placing more orders per day or designing the optimal bake plan per day requires factoring in both weekday-related and within-the-day variation in demand. For some products, the within the day or so-called intraday demand pattern will follow the general customer footfall for that day; for other products, such as lunch items, demand is more influenced by how the items are planned to be consumed.

Demand influences

Figure 11: For some products, such as the laundry powder above in green, the intraday demand pattern will follow the general customer footfall pattern. For other products, such as lunch items, demand is more influenced by how the items are planned to be consumed.

Again, keeping track of both weekday and intraday demand patterns manually is quite a complex and error-prone process. Yet, many retailers still rely on their store associates to figure this out on their own. This is a high-stakes gamble, as ultra-fresh products inevitably have a big impact on how consumers judge the quality of fresh products in a store.

Best-in-class retail planning systems can figure out the optimal split between multiple orders or production batches per day as well as adjust the quantities as needed, automatically.

3.1.4. Adding Science to the Art of Managing Fruits and Vegetables

Fruits and vegetables are often last in line when store ordering is automated. Obviously, produce faces the same challenges caused by short shelf life and variable demand as other fresh product categories.

In addition, the varying supply and quality of fruits and vegetables demand additional flexibility from the planning system in use.

The regions from which fruits and vegetables are sourced constantly change as crops are harvested in different parts of the world at different times. Even growers in the same region may have timed their crops slightly differently. Furthermore, as there is always some uncertainty in the availability of good quality product, food retailers usually try to ensure that they always have several vendors for the same product.

From a consumer perspective, a lemon is a lemon, but the supply chain may need to deal with tens of different product codes for lemon, each associated with a different vendor. Effective management of fruits and vegetables requires the planning system to be able to seamlessly switch between planning levels as needed:

  1. 1. Demand forecasting needs to be conducted on the product level, for example “domestic organic tomato,” using historical sales data for all domestic organic tomatoes regardless of supplier.
  2. 2. The replenishment quantity also has to be determined based on the available inventory of domestic organic tomatoes as well as their forecasted demand.
  3. 3. The replenishment order, however, should be generated for the current vendor of domestic organic tomatoes. This is where the planning system needs to move from working on the product level to the SKU level, i.e. from “domestic organic tomato” to “domestic organic tomato supplied by GreenGrowers Co.”
  4. 4. Often, the replenishment order needs to be split between two or three vendors to ensure availability in case of vendor product shortage as well as to keep several vendors in business. In that case, the planning system also needs to take care of allocating the order need to several vendors — for example 65% to GreenGrowers Co and 35% to OrganicFarmers Co.

The forecasting and replenishment process for fruits and vegetables is highly laborious to manage manually but can be effectively automated. The key prerequisite is clear guidelines for which products are to be included in the stores’ assortments and which vendors are to be used for sourcing at any given time. As in any automation process, high-quality master data is essential.

3.2. Optimized Center Store Replenishment Is Key to Supply Chain Efficiency

Fresh products need to be delivered to stores in perfect sync with demand. Center store products and other products with longer shelf-lives, on the other hand, offer more opportunities for an optimized flow of inventory in the supply chain. Optimized replenishment of center store products is key to lowering costs in stores and throughout the grocery supply chain.

3.2.1. Smart Replenishment for Efficient In-store Goods Handling and More Level Goods Flows

Typically, every large grocery retailer replenishes all or at least most of its stores every day from its distribution centers. This is because fresh products demand frequent deliveries and because the overall inventory flows are substantial enough to warrant daily deliveries.

If all replenishment opportunities are used for all product groups without discretion, two problems will follow:

  1. 1. The deliveries to the stores will consist of a random mix of products from several product categories displayed in different parts of the store. This means that store personnel will spend a significant amount of time moving roll cages around the store to stock shelves (see Figure 12).
  2. 2. The delivery volumes on different weekdays will not be roughly equal, but rather will reflect the daily variation in sales volume, often with significant delivery peaks towards the end of the week in anticipation of weekend demand. This leads to fluctuating capacity needs in both distribution and stores, which increases costs.

Replenishment days

Figure 12: When roll cages contain a large variety of products, store personnel spend a lot of time moving from aisle to aisle while shelving products. The use of main replenishment days based on the store’s floor plan significantly increases shelving efficiency.

Instead of automatically using all available order or replenishment opportunities for all products, the best practice is to define main replenishment days for longer shelf life products. This means that replenishment of some center store product groups is concentrated to specific weekdays instead of being scattered throughout the week. Replenishment planning, such as the optimization of safety stock and calculation of order quantities, will be based on delivering the goods on the specified main replenishment days. However, to ensure the highest possible availability, replenishment orders are also triggered for the other available replenishment days to avoid stock-outs if there are unexpected demand peaks.

In practice, this means that instead of ordering detergents every day, fast moving detergents are primarily replenished on, for example, Mondays and Thursdays, and slow-moving detergents on Thursdays. For detergents, the other replenishment days from the distribution center to the store are only used in case there is a risk of stock-out in the store.

The use of main replenishment days allows for significantly more efficient in-store replenishment without hampering on-shelf availability. More consolidated deliveries make it more efficient for store personnel to replenish store shelves, especially when the main replenishment days are set based on what product categories are displayed in the same aisle or zone of a store. We have seen reductions of 20% in the time spent stacking shelves following the introduction of main replenishment days.

Linking stores and dc's

In addition to creating more consolidated deliveries, main replenishment days enable leveling out inventory volumes between weekdays. In many stores, weekends can be very busy, with lots of customers doing their weekly shopping while large quantities of fresh products are being delivered to the stores. Setting main replenishment days for center store products to the quieter weekdays balances the incoming goods flow and makes personnel planning in the stores easier.

As with many other processes, the use of main replenishment days can be further optimized when the basics are in place. For stores that have higher footfall on weekends, additional capacity management may be beneficial. If several main replenishment days are in use, the one closest to the weekend tends to get most of the volume. To further level out the inventory flow, best-in-class retail planning systems can look at projected orders for an upcoming week, identify if there are undesired peaks in inventory volumes, and automatically move some of the replenishment volumes to quieter days.

3.2.2. Space-aware Replenishment for Efficient Goods Handling

Food retailers have traditionally operated in a very siloed manner with very little communication between the merchandizing teams responsible for store planograms, the supply chain teams responsible for store replenishment, and the store operations teams responsible for in-store work processes. This must change.

The space allocated to each product in a store has a big impact on both the results and costs of the store replenishment process:

  • If the allocated space is very large in comparison to demand, the inventory needed for ensuring optimal on-shelf availability will not be sufficient for maintaining a visually appealing, full display. For that purpose, additional visual minimums need to be defined. Visual minimums indicate how many units of a product need to be on the shelf to ensure that the display is visually appealing. For slow-sellers, the visual minimums will always be higher than the inventory levels required for great on-shelf availability. For long shelf life products this may not be a problem, but for fresh products, excessive visual minimums may cause unnecessary spoilage.
  • If the allocated space is small in comparison to demand, incoming deliveries will not fit on the shelf. At least part of the delivered quantity will need to be placed in a backroom or other storage area. This significantly increases the cost of shelf stacking, as goods need to be moved back and forth between the sales area and the backroom. In addition, the use of backroom storage significantly increases the risk of empty shelves, as timely replenishment from the backroom is dependent on the vigilance of store personnel.

Although surprisingly rare, full integration between space and replenishment planning is an important best practice for increased operational efficiency:

  • Access to planogram data makes it easy to automate the maintenance of visual minimums on the product-store level based on the number of facings or total shelf space allocated to each product in each store.
  • Access to planogram data makes it easy to automatically trim replenishment orders that would cause incoming deliveries not to fit on the shelf. Usually this rule needs to be balanced with the risk of stock-outs if the space allocated to some products is very small in relation to their demand.
  • Access to floor plan information enables assigning main replenishment days based on where in the store products are displayed, with the aim of creating more focused deliveries that minimize the need for store personnel to unnecessarily move around the store when stacking shelves.
  • Access to planogram information makes it possible to plan replenishment so that shelves are filled up to the maximum each time a delivery comes in, minimizing shelving work in stores. This means that rather than getting two batches in one go, if there is space for a third one that would be delivered next week, the order is calculated to fill the assigned shelf space upon arrival.
Visual minimums

The space assigned to each product is of vital importance to how efficiently the replenishment process can fucntion, so it is important to deliver continuous feedback to merchandizing. Good analytics tools will help you identify products and stores where there is a mismatch between space and sales, i.e. products and stores for which incoming deliveries do not fit directly on the shelves or products and stores where visual minimums lead to waste or markdowns.

Ideally, space planning should always be based on the detailed store, product and day-level forecasts as well as information on replenishment cycles and main replenishment days available from replenishment planning:

  • By using the accurate forecasts rather than looking at historical sales data when optimizing how space is allocated to products, it is much easier for the space planning team to take seasonality and trends into account.
  • Based on good forecasts of the expected maximum sales per delivery interval, shelf space can be optimized to be truly efficient for all products in a store on all weekdays. This kind of optimization makes it possible to attain fewer deliveries and direct-to-shelf flows for a much bigger proportion of the product range.

We have seen forecast-based optimization of shelf space translate into up to 30% lower distribution and in-store replenishment costs.

3.2.3. Dynamic Pack Sizes to Meet Dynamic Demand

One powerful tool to increase store replenishment efficiency is to optimize the use of different pack sizes. In many cases, stores can choose to order case packs, pallet layers or full pallets from the distribution center. Larger batches are more efficient to handle both in the stores and at the distribution centers, but clearly the deliveries need to match the available space and demand in the stores. Otherwise inventory will pile up in the stores and reduce rather than increase efficiency by congesting back rooms and causing multiple trips between the backroom and shop floor to replenish shelves.

Especially for retailers operating stores of different sizes, optimizing replenishment pack sizes per product and store has a direct impact on handling costs. However, doing it only once as a concerted effort does not suffice as demand changes over time and, for some products, also with the seasons. During the high season, a pallet might be most efficient while outside the peak, smaller case packs may be more efficient.

The retail planning system needs to be able to automatically optimize which pack size to use per product, store and order. This means that whenever an order is placed, the system always checks all available pack sizes—typically varying from the case pack to full pallets—and selects the most efficient pack size in relation to forecasted demand.

The attain the full efficiency gain, the supplying warehouses need to be able to estimate the demand for the different pack sizes. Otherwise they may end up in a situation where they use individual packs to put together pallets for the stores, rather than having full pallets flow through the distribution system. This is possible when the store projections (see Section 6.1) used as the basis for distribution planning reflect the stores’ forecasted use of different pack sizes.

4. An Integrated Supply Chain Driven by Customer Demand

Traditionally, store replenishment and inventory management at the regional distribution centers or central warehouses have been separate processes, driven by separate demand forecasts.

In a 2018 survey, we found that 16% of large US grocery retailers still base their distribution center forecasts on historical data of outbound deliveries from these distribution centers. This is akin to driving a car while looking at the rearview mirror.

Forecast accuracy

According to the same survey, a clear majority of retailers—70% of the respondents—have chosen the more forward-looking approach of basing their distribution center forecasts on store demand forecasts. Granted, this is a better approach than only looking at outbound deliveries.

There are, however, some important disadvantages to using store demand forecasts to drive planning at the distribution centers:

  1. 1. Goods need to be delivered to the stores before the stores can sell them. This means that the distribution center forecast needs to go up before the stores’ demand forecasts go up and vice versa. The difference in timing depends on the stores’ sell-through rates and replenishment schedules, which means that the difference in timing varies between stores and products and sometimes also weekdays. The result is that it is almost impossible to accurately account for the difference in timing, which is bad news for your forecast accuracy at the distribution centers.
  2. 2. When goods are pushed rather than pulled through the supply chain, there will be outbound delivery peaks at the distribution centers which are not visible in the stores’ demand forecasts. A typical example is promotions, where anything between 30% and 100% of the expected promotional uplift needs to be delivered to the stores before the promotion starts. The promotion, thus, causes a much bigger demand peak at the distribution center than in the stores. This peak is fully controlled by the retailer itself, but still requires a lot of manual planning work or “guesstimation” when the supply planners at the distribution center try to anticipate when and in which quantities stores will take in the promoted products.

It is quite ironic that many of the situations considered most difficult to tackle in the distribution centers, such as building up stock in stores for promotions or new product introductions, are situations fully in the hands of the retailers themselves.

The best practice is to base distribution center forecasting on the stores’ projected orders, which reflect both pull-based demand as well as planned, push-based stock movements. In a 2018 survey, only 14% of the responding North American grocery retailers had implemented this.

To achieve seamless integration of store and distribution planning, the planning system needs to be able to calculate projected store orders per product, store and day several months or even a year into the future, reflecting current and known future replenishment parameters as well as the demand forecast. These calculations, of course, require significant data processing capacity, which is likely to be one explanation for the surprisingly low adoption rates.

Integrated supply chain

Figure 13: An integrated supply chain is driven by consumer demand, taking all known factors such as delivery schedules, on-hand inventory and pack sizes into account. The shipment plan for the distribution centers consists of projected store orders as well as demand forecasts for potential direct-to-customer inventory flows, such as online orders picked at the distribution center.

In practice, the stores’ order projections consolidate data on their current inventory, safety stocks and visual minimums, delivery schedules (including main replenishment days) as well as any planned inventory movements, including everything from stocking up to build promotional displays to shifting orders to level out the capacity requirements in distribution.

Table 1 presents some examples of situations in which the value of basing forecasting at the distribution centers on projected store orders is especially notable.

When the order projections are aggregated across all stores, they form a very accurate, customer-driven forecast for the distribution centers.

Additional benefits of the supply chain integration include supply chain transparency supporting capacity planning, supplier collaboration (discussed in Section 6.4) as well as straightforward handling of cross-docking, pick-to-zero, and shortage situations.

Product introductions When launching a new product, at least one case pack or a sufficient amount of product to fill up the allocated shelf space is pushed out to each store. This creates inventory buffers in the stores, which will take days or weeks to digest. As long as there is surplus inventory in the stores, the projected store orders (as well as the actual outflow from the distribution centers) will be lower than forecasted consumer demand.
Product terminations When a product termination has been planned in advance, the distribution center forecast will automatically go down as the termination date draws closer, supporting a controlled ramp-down of inventory. When the distribution center forecast is based on projected store orders, the forecast automatically considers the existing inventory buffers in the stores and accurately estimates how long it will take to clear out the remaining stock in each store.
Promotions Typically, anywhere between 30 – 100% of the expected promotional uplift is pushed to the stores in advance of a promotion. The good news is that these planned inventory movements are completely predictable (as they are in fact planned, with no need for forecasting) and will automatically be included in the projected store orders. Also, if the stores are left under- or overstocked following the promotion, the stores’ fulfilment needs will be accurately reflected in the distribution center forecasts.
Seasons Almost always, some buffer stock is distributed to the stores before the start of a major season. This can be due to the need to create nice seasonal displays in the stores, level out seasonal peak volumes, or due to the season being weather-driven making the exact timing of the season start somewhat uncertain. As with promotions, these planned stock movements will be automatically visible in the stores’ projected orders used as the forecast for the distribution centers. Furthermore, as seasonal demand may vary a lot between stores, for example due to local weather conditions, the stores’ inventory buffers will be consumed at different paces. This will be automatically visible in the forecast for the distribution center.
Changes in replenishment schedules It is not uncommon that stores replenishment schedules are changed either temporarily, for example to match increased demand in the high season, or permanently, for example following the implementation of new transportation routes. Changes in the replenishment schedule will, naturally, not have an impact on consumer demand but they will have a direct impact on the goods flow into the stores. The resulting changes in the timing and size of the deliveries to the stores will automatically be captured in the distribution center forecast when it is based on projected store orders.

Table 1: Examples of situations where the use of projected store orders, rather than stores’ demand forecasts, allows for much more accurate planning at the distribution centers.

4.1. Plan Once and Execute Automatically Throughout the Supply Chain

When planning at the distribution centers is based on the stores’ projected orders, the impact of planned activities, such as promotions or pre-season allocations, are immediately visible throughout the entire supply chain. To reap the full benefits of this transparency, all planning data needs to be made available to the planning system as soon as a promotion plan, assortment change, price change, or any other relevant decision has been made.

A planning system that supports time-dependent master data is a key enabler of proactive planning. Below are just a few examples of how time-dependent master data enables you to register valuable information immediately when it becomes available. This, in turn, allows your replenishment planners to rely on the planning system to automatically trigger the necessary actions at the right time with very little manual work.

  • Time-dependent replenishment schedules: When store replenishment schedules can be managed using dates, it becomes possible to update the planned future replenishment schedules into your planning system as soon as the information becomes available. This enables replenishment planners to trust the planning system to automatically consider these changes both in replenishment planning and when calculating supply chain projections.
  • Assortment activation and termination dates: When start and end dates for the active product range have been defined, product ramp-ups and ramp-downs are much easier to manage. Routine planning tasks, such as pipeline fills for new products or inventory ramp-downs for products to be discontinued, can be automated. This automation reduces manual work, but also ensures optimal inventory levels in all phases of a product’s lifecycle.
  • Stock-ups before promotions: Promotions naturally have start and end dates, but it is equally important to be able to specify beforehand how stores should be stocked. It is usually ideal to define how many days before the promotion the promotional goods should arrive in the stores, what stock quantities stores should receive to be able to build the planned promotional displays, and what proportion of the forecasted promotional demand the first deliveries should cover. Rules and templates make it possible to attain accurate replenishment plans for each store and product without manual work.
  • Temporary supplier delivery restrictions: Suppliers may have temporary delivery restrictions. Chinese manufacturers may, for example, not dispatch shipments during the Chinese New Year. If information like this is made available to the planning system, the system knows to put in orders for this period early enough to ensure high availability during the affected period, while minimizing manual work and dependence on human memory.

An integrated supply chain set-up removes the need for double-planning work. The impacts of planned changes in store replenishment are automatically reflected in the projected store orders forming the demand forecast for the distribution centers. This means that as soon as the required store stock-ups for promotions are planned, they will be visible in the distribution center forecast on the right dates and in the right quantities.

Of course, having the right functionality in your planning system is a key enabler, but the real challenge is getting the whole organization to work more proactively. Ensuring that decisions are made early enough, but not too early to unnecessarily reduce flexibility in a dynamic market, requires that everyone in the organization has a basic understanding of how the supply chain works and what the relevant lead times for different types of decisions are.

4.2. Multi-echelon Optimization of Goods Flows

An integrated supply chain makes it possible to manage multi-echelon inventory flows efficiently, with minimum waste and a high level of automation. When all data on demand forecasts, available stock, delivery schedules, lead times and batch sizes for all supply chain echelons is available in the same planning system, it enables seamless optimization of inventory flows throughout the supply chain.

Cross-docking is an inventory strategy aimed at maximizing transportation efficiency while minimizing handling costs. Cross-docking is often applied to bulky products, such as drinks, to attain lower storage and handling costs. It can also be used to cut lead-times for short shelf life products. In a cross-docking set-up, goods are delivered from the supplier to a cross-docking facility where the goods are put not into storage, but moved from the inbound truck to an outbound truck for distribution to stores.

There are some requirements for cross-docking to work efficiently: 1) Suppliers need to be able to deliver full truckloads to the cross-docking facilities, 2) the delivery units, such as pallets or roll-cages, need to be ready for immediate movement to the outbound trucks without additional handling, and 3) the outbound trucks need to get a high fill-rate to keep transportation costs down. The planning system, thus, needs to optimize both inbound and outbound flows to and from the cross-docking facilities as well as understand the total lead time from supplier to store.

Another example of an inventory policy that requires integrated supply chain planning is pick-to-zero. In this inventory strategy, orders to the suppliers are based on the stores’ replenishment needs. However, rather than fixing the quantities to be delivered to each store, the supplier delivery is reallocated to the stores upon receipt based on the latest inventory and forecast information. This allows for adjusting the delivery quantities per store in case the supplier could not deliver in full or in response to potential unexpected demand peaks in the stores after the original replenishment need was calculated. As a result, supply matches demand more accurately than when using the traditional cross-docking approach. The pick-to-zero approach can be seen as a way of shortening the order-to-delivery lead times to the stores, as the store-specific quantities are finalized not when ordering from the suppliers but when preparing the goods for store distribution.

When supply chain planning is fully integrated, exceptions can be resolved in an optimal and automated fashion. Let’s look at inventory scarcity due to, for example, an incoming shipment being delayed. Instead of fulfilling store orders on a first-come, first-served basis, the available inventory can be automatically allocated to stores to maximize overall on-shelf availability or in accordance with a tactical prioritization of the stores. In the best case, on-shelf availability is not even affected. In a similar manner, inventory batches nearing their expiration dates can proactively be forced out to the stores that have the best chance of selling the products at full price.

5. Efficient Inventory Management in Distribution Centers

Replenishment of central warehouses and distribution centers is sometimes seen as more of an art than a science. It is true that longer lead times, especially when ordering overseas, and lack of control over external suppliers introduce complexities. Yet, at least in principle, replenishing central warehouses or distribution centers is not that different from replenishing stores.

When replenishing stores from their own distribution centers, retailers can optimize order fulfillment as they find best. When ordering goods from suppliers, though, there may be complex restrictions regarding minimum order value or quantity. In addition, there may be volume-based discounts or other rebates which, when efficiently exploited, can have a significant impact on margins. Many retailers have not been able to put this kind of supplier contract or price information into their planning systems, making it necessary for the operative buyers to invest significant time double checking orders.

When replenishing stores, the active goods flows (combinations of products and stores) for any larger retailer is typically measured in millions or tens of millions, which means that automation is crucial. For central and regional warehouses, the number of order lines is much smaller and the value per order line much higher, making the economic impact of each order line more pronounced. This has both enabled and encouraged a lower degree of automation in operative buying compared to store replenishment.

We have found that setting up operative buying processes in a structured way with good system support can also result in very high levels of automation at the distribution centers. This does not, however, necessarily mean that best practice retailers have a significantly leaner buying staff. A key result of increasing the automation of routine tasks is that operative buyers have more time to proactively deal with potential capacity, delivery or quality issues and to analyze the performance of the current assortment, suppliers and supplier agreements for continuous improvements.

5.1. Total Cost Optimization of Inbound Flows

As the inbound goods flows to distribution centers are more consolidated than the outbound flows, there are more opportunities for order optimization when replenishing distribution centers than when replenishing stores.

It is important that the planning system can perform order optimization on multiple levels to reach the most cost-efficient outcome.

Some examples of order optimization on different levels are:

  • Calculation of the economic order quantity (EOQ) per product to minimize inventory and handling costs
  • Selecting the optimal order batch size—such as case pack, pallet layer or full pallet—when several order batch options are available, considering potential price differences between the different batch options
  • Building mixed pallets for efficient transportation and goods handling
  • Building orders that fill one or multiple load carriers, such as trucks or containers, or that meet suppliers’ order restrictions, such as minimum order value or minimum number of pallets

Although it seems simple, the process of pooling orders for multiple products to fill load carriers or meet supplier order limits can be quite the test for your planning system’s flexibility.

To meet supplier requirements and benefit from lower transportation costs or supplier discounts without accumulating excess stock, you typically need to be able to:

efficient ordering and handling
  • Flexibly define which products should be pooled when planning an order. Products from the same supplier are often pooled together, but sometimes the same supplier’s different manufacturing sites should be considered separately or all products sourced from the same region, regardless of supplier, seen as one group.
  • Set targets and/or limits for the consolidated order in multiple units, such as value, volume, number of pallets, weight or combinations of these dimensions. For example, when filling trucks, you want the order to fill the available space efficiently so as not to incur costs for transporting air, while at the same time ensuring that the legal maximum weight is not exceeded.
  • Let the planning system decide which kind of a load carrier it should aim to fill up with the order. With some suppliers, it may make sense to sometimes order a truck, sometimes order a truck and a trailer, and sometimes two trucks, only one of which has a trailer, depending on the forecasted demand.
  • Set the right order trigger level. When supplier order restrictions are difficult to meet, it may make sense to require enough demand for, say, at least 30% of a truckload before the planning system starts building an order that fills up the entire load carrier.

In addition to letting the planning system do the heavy lifting when it comes to supplier order requirements, the best practice is to constantly evaluate these restrictions and their impact on the flow of goods. Multi-year contracts in a dynamic market or fixed order restrictions for products with seasonal demand may turn out to be costly or infeasible as demand changes.

To support this, the ideal planning system should highlight all order suggestions more/less than needed as a result of these constraints, as well as show the difference from the actual need. Furthermore, it should provide analytical support to help the operative buyers make rational decisions concerning whether the benefit, such as rebate, of meeting a supplier restriction is greater than the resulting increase in inventory carrying cost and risk of obsolescence.

5.2. Smart Buying Takes Advantage of Good Prices

Retail costs are dominated by the cost of goods sold. The operative buying team needs to take responsibility for efficiently exploiting rebates to improve gross margins.

In theory, smart buying when prices are changing is quite straightforward:

  • When you know that the price of a product will go up, stock up just before the price increase.
  • When you know the price of a product will go down, only order the quantity you absolutely need before the price change, and then stock up after the new price has come into effect.
  • When a price will be lowered temporarily, for example due to a supplier campaign, order less just before the price reduction and stock up when the price is low.

To be able to truly benefit from price changes, you also need to factor in your inventory carrying cost, time your orders correctly relative to when the price is changing, and potentially split the investment buy—the additional quantity you are buying above what you would need to meet demand—into several orders.

To further complicate things, there may be other factors that have an impact on the optimal order quantity. For fresh products, shelf life is always a factor. It makes absolutely no sense to stockpile inventory that will end up as waste, or to harm your reputation by putting goods with unattractive expiration dates in stores. Furthermore, in situations where storage space is scarce, the cost of inventory may suddenly jump to a whole new level if you exceed the capacity limits of your current warehouses. When your storage space is very full, you would need to find additional space outside of your current warehouses for additional goods, quickly turning your investment buy into a very unprofitable move.

The best practice is to feed your planning system with time-dependent price data to let the system optimize when and in what quantities to buy when prices are changing. In this way, you can take advantage of even minor price changes effectively, as the operative buyers do not need to spend time manually figuring out the optimal order quantities. It is important to keep in mind that restrictions, such as shelf life for perishable items or capacity limits on storage space, need to be considered. If your planning system is not able to deal with such restrictions automatically, the suggested investment buys will need to be double checked by the buying team.

It is not unusual to have supplier contracts include a rebate triggered by the buyer’s annual order value exceeding a set quota. Again, keeping track of supplier quotas, placed orders, and forecasted orders is very hard to do manually. Intelligent planning systems support smart buying decisions by suggesting additional orders to get the rebate when feasible and by not suggesting any additional orders that would result in counterproductive stock-piling.

5.3. Batch Level Inventory Management of Perishables

It is currently impossible to know the exact expiration dates of on-hand inventory in stores. It can even be hard to get a decent estimate if there are several batches of a product simultaneously on the shop floor, as some consumers work hard to find the freshest products available.

However, distribution centers are a different story. In distribution centers, modern warehouse management systems ensure that inventory is shipped on a first-in-first-out basis. In addition, they keep track of the exact expiration date for each batch in stock.

Making good use of batch-level expiration data in inventory management reduces waste and improves your service level:

  • When the planning system is able to calculate projected spoilage based on forecasted demand and the expiration dates of the available inventory, it can order a sufficient amount of goods to replace soon-to-expire inventory before the products expire. This significantly improves the service level of your distribution centers towards your stores.
  • Best-in-class planning systems trigger early warnings when there is a risk of waste. This enables efficient force-outs to those stores most likely to be able to sell the products at a good price, or it gives you time to find other sales channels, such as discounters, which may be willing to take the batches off your hands at a discount.
  • The spoilage projections also show if there is a systematic risk of waste, for example due to products having large order batches compared to demand.

Figure 14: High quality spoilage projections clearly show if there is a systematic risk of waste, for example due to too large an order batch size of a product.

5.4. Real-time Data for Buying Fresh Products

For perishable products, a very high inventory turnover both in stores and in the supplying distribution centers is a must. This means that the supply chain is very vulnerable to quality issues, delivery problems or unexpected peaks in demand. In situations where store requirements exceed available inventory, quick reactions are of essence.

In many cases, suppliers of short shelf life perishables make several daily deliveries to the same distribution centers. This is partly to guarantee freshness and partly to level out volumes.

Several daily supplier deliveries make it possible for a retailer to accommodate actual demand by placing the orders as close to the different ordering deadlines as possible, making use of the latest demand and inventory data.

However, to be able to identify demand surges, the planning system needs to be tightly coupled with the underlying transaction systems and have access to real-time data. Of course, the planning system also needs to be able to process the data quickly enough to turn the latest data into orders as accurately as possible.

Similar quick reactions and within-the-day calculations based on real-time data are valuable when fruits and vegetables, which are prone to supply and quality issues, are received in the morning. Because the actual available inventory may differ from what is planned, it makes sense to re-allocate stock based on the latest forecast and stock data from the stores rather than fulfilling store orders in an arbitrary order.

6. Planning for Optimal Capacity Utilization in the Retail Supply Chain

In a dynamic business like retail, capacity bottlenecks can emerge in almost any part of the supply chain in response to a range of events — for example holidays, unusual weather, big promotions or big assortment updates in the stores.

Major holidays like Christmas require a massive amount of inventory be pushed through a supply chain optimized for a much smaller flow of goods. This requires meticulous planning several months ahead of the season. Mistakes in capacity planning result in products not being available for sale until after the season or staff needing to put in excessive overtime on short notice. Retailers’ reputations as well as millions of dollars are at stake.

Despite its importance, capacity planning at many food retailers is still a very manual, spreadsheet-based process. Even in the hands of seasoned professionals, this can be considered a high stakes game of “guesstimation,” where the cost of planning errors can quickly amount to millions. Furthermore, a highly manual, spreadsheet-based approach turns capacity planning into a massive endeavor that can only be conducted for the biggest events. This means that capacity bottlenecks that could have been foreseen and proactively addressed routinely go undetected, causing unnecessary costs and fire-fighting in the supply chain.

The best practice is to use high-quality supply chain projections to plan capacity and foresee bottlenecks in all parts of the supply chain. Leading retailers use this planning data throughout their organizations to support not only inventory planning but also capacity and workforce optimization both in distribution centers and in stores. They also help their suppliers prepare by sharing planning data with them in a collaborative fashion.

6.1. Supply Chain Projections Form the Foundation of Capacity Planning

Supply chain projections are essentially the result of a detailed simulation of future goods flows throughout the supply chain. As explained in Section 4, these supply chain projections capture both 1) demand-driven goods flows, mainly replenishment orders triggered to meet forecasted demand in the stores and distribution centers and 2) planned stock movements for building up stock in preparation for promotions or ramping up new products. In addition, high-quality supply chain projections consider time-dependent master data, such as planned future store replenishment schedules.

Proactive planning

Supply chain projections calculated on the SKU and day level for the entire supply chain enable retailers to understand the capacity requirements in any part of the supply chain several months or even a year ahead. The same projections allow you to foresee the need for transportation between different nodes of the supply chain, the required storage capacity in both stores and warehouses, the amount of order lines in each distribution center that need to be picked on specific days, and so on.

Figure 15 below demonstrates how projected inventory balances for products stored in a specific storage area, such as the frozen goods warehouse, can be aggregated to uncover the future need for frozen storage space and to foresee potential capacity bottlenecks in frozen storage.

Projected inventory

Figure 15: By aggregating projected SKU level inventories for a specific warehouse area, a retailer can very accurately foresee its future need for storage space and act in case there is a risk of capacity limits being exceeded.

The best practice is to not only calculate supply chain projections, but to also set up automated exception alerts and, when appropriate, automated exception management routines to deal with situations where projections show that capacity limits will be exceeded. Examples of capacity-based exceptions include the risk of exceeding storage capacity, the risk of exceeding order picking capacity, and the risk for spoilage due to excess stock of perishable items.

When projections are calculated sufficiently far into the future, they provide the necessary support to move away from spreadsheet-based ad hoc modeling in preparation of major holidays toward truly data-based planning of key events.

6.2. Retail Sales & Operations Planning for Increased Profitability

We use the term Retail Sales & Operations Planning (Retail S&OP) for the process of balancing supply chain capacity and resources with consumer demand in the medium or longer term. As discussed in Section 3.2.1, retailers’ supply chain teams routinely balance the need for capacity, for example by leveling out the flow of goods between different weekdays. However, the Retail S&OP process is typically a more cross-functional effort with representatives from the supply chain, store operations, logistics and warehousing, merchandizing, and suppliers working together to find solutions that benefit all parties.

An important use case for Retail S&OP is preparing for major holidays, Christmas being the most important example in most markets. During those holiday seasons, demand is typically much higher than normal, with demand usually increasing steeply until the holiday peak. After the peak, demand drops back to normal or even drops to a lower than usual level for a while.

The target of the Retail S&OP process is not just supply chain efficiency; it is maximum profitability. The Retail S&OP should result in:

  1. 1. Goods flows that do not exceed capacity restrictions in any part of the supply chain. This secures a reliable supply, which in turn ensures sales are not lost due to delivery problems.
  2. 2. Cost-effective operations by minimizing the need for costly overtime in all parts of the supply chain.
  3. 3. Informed decisions in the interest of total company profitability in situations where it is impossible to guarantee perfect on-shelf availability of all products due to capacity restrictions.
  4. 4. Transparency into resource requirements, making it possible to ensure that all resources, including store workforce availability, are based on the same operational plan.

The process of preparing for major holidays typically starts around six months prior to the season in question. The first step is to agree on the constraints: Will there, for example, be changes in delivery schedules/lead times or supplier capacity constraints due to the holiday season?

The next step is to ensure that sales and delivery plans are reviewed in order to identify potential bottlenecks. These bottlenecks can appear in any part of the supply chain in different phases of the holiday season. For example, here may be too-large deliveries congesting the stores on some days; more order lines than the picking automation in the retailer’s warehouse can handle on a specific day; more frozen products to be stored than there is available space in the warehouses; and so on.

The only way to detect moving capacity bottlenecks with a reasonable degree of certainty is to use supply chain projections. Identifying bottlenecks by modeling the supply chain with all its complexities is impossible using spreadsheets. Building a slightly simplified model will still be both very time-consuming and highly error-prone.

When the potential bottlenecks have been identified, actions to remove them need to be examined using what-if scenario planning. Bottom-up scenario planning allows a retailer to see exactly how changes in the timing of deliveries, replenishment schedules, or forecasted sales volumes would impact on the goods flow throughout the supply chain.

Typically, food retailers need to deliver products with longer shelf-lives to the stores earlier in order to free up capacity for dealing with fresh products in the high season. Tools for leveling out the goods flow include:

  • Filling up shelves in stores: For many center store products, full shelves cover the demand of many weeks. If waste is not an issue, it makes sense to fill those shelves ahead of time in November so that all of December can be used to deliver fresher products.
  • Allocating longer shelf life products to stores: Fill-ups respect the assigned shelf space in the stores, but that isn’t always enough. Sometimes, it makes sense to allocate the demand for the next 2-4 weeks to stores in one go. Those items will then be stored in the backroom of a store and shelved from there when needed. This is not suitable for all products but can be useful for special cases.
  • Making changes to delivery schedules: Most food retailers already have frequent deliveries to their stores outside of the holiday season, but in some cases, it may make sense to add deliveries to compensate for the increased demand.

When a satisfactory scenario has been identified and agreed upon with suppliers, the retailer can lock their plan well in advance of the season and focus only on execution and corrective actions from there on out.

An effective S&OP process leads to more level capacity utilization throughout the season. In addition, whatever peaks remain are known beforehand rather than popping up late as costly surprises.

Improved capacity planning

Figure 16: Example of the effect of improved capacity planning for major holidays. After the introduction of a new Retail S&OP built around supply chain projections, this retailer has successfully managed to level out the need for capacity. The remaining peaks are also known well in advance. The attained savings account to several millions of dollars just for the Christmas season.

6.3. Workforce Optimization for Improved Service at Lower Cost

Another area where medium to long term planning is highly valuable is workforce planning and optimization. In many countries, legislation dictates that work shifts need to be announced several weeks or even months ahead of time, making it important to be able to create accurate rosters well in advance. Even when not required by law, retailers can increase employee satisfaction by providing information on future work shifts reasonably early.

In traditional brick-and-mortar grocery retail, store personnel form the largest operational cost, amounting to approximately 14% of revenue.

The need for workforce in stores is mainly driven by two factors: 1) Incoming stock that needs to be put on shelves and 2) the number of customers that need to be served, mainly at check-outs. Also, 3) if online orders are fulfilled from stores, this needs to be factored in, as it is typically quite labor-intensive and time-consuming.

Workforce optimization

In practice, the following forecasts are required for efficient workforce optimization:

  • Forecasted customer footfall on hourly or 15-minutes intervals, which is the driver for check-out work and customer service.
  • Forecasted incoming delivery volumes per product type, such as fruits and vegetables, frozen etc. per day, which is the driver for in-store goods handling work, such as shelf stacking.
  • Forecasted online order lines to be fulfilled within a certain time frame (depending on the offered delivery lead times), which is the driver for order picking and packing work.

Accurate workload forecasts in combination with automatic optimization of work schedules typically enable a 10 – 15% reduction in store personnel costs, while at the same time supporting better customer service.

Though it is an innovative approach, workload forecasting and optimization is actually surprisingly easy to implement. Retailers that have already implemented best practice demand forecasting (see Section 2), will not face any significant challenges in producing the footfall or online order forecasts. Furthermore, forecasted incoming delivery volumes are automatically available when high-quality store order and delivery projections (see Section 6.1) are available in the planning system.

Workforce-optimization

Figure 17: When rosters are planned manually, there is typically both over- and under-capacity in the same stores at different weekdays or times. With a more data-driven approach to workforce planning and optimization, rosters meet the actual need for staff more accurately, resulting in a more level workload and reduced personnel costs.

The roster planning at warehouses and distribution follows a similar pattern, with the main workload drivers being 1) projected outbound delivery lines and 2) projected incoming delivery lines.

6.4. Efficient Supplier Collaboration

Supplier collaboration has been discussed for decades, but surprisingly few retailers have implemented it extensively. To establish a fruitful collaboration process, both parties need to put in some effort and attain measurable benefits. This has rarely been the case, which is why many collaboration initiatives have failed.

Technology does not solve this challenge, but it can ease the pain. Typically, a majority of the work in collaboration programs has been spent on collecting data from various sources. With the right planning system in place, this work can be minimized. We also recommend building your supplier collaboration processes bit by bit rather than trying to fix everything in one go.

A good starting point is to share your order forecasts with your suppliers. For retailers, this is a very lean way to collaborate. If you have already invested in a planning system with the capability to calculate supply chain projections, the purchase order forecast, which tells your supplier what you are planning to buy from them in the weeks and months to come, is already readily available. You only need to set up automated reports to routinely share this information with your suppliers.

Supplier collaboration

The purchase order forecast can be further complemented with information on planned promotions and other events or changes, helping the supplier understand the reasoning behind your purchase order forecast. You can also share demand forecasts or point-of-sale (POS) data, but the most essential information to share is what you expect the supplier to deliver and when.

A more collaborative way of working requires that both parties be willing to invest more time and effort and that both parties see the value this will bring. This sort of collaborative planning, often referred to as CPFR (Collaborative Planning, Forecasting and Replenishment) is truly two-way communication, whereas forecast sharing is mainly one-way communication. Again, a best-in-class planning system helps by providing reliable projections of future purchase orders, analytical tools to understand potential changes and issues, as well as a platform or portal for the collaboration.

Optimally, suppliers can log into a portal to see the retailers’ view on how their items will sell and how the retailer is planning to place purchase orders, coupled with data on promotions, seasons, events and so on. Suppliers can then add their own view. The supplier’s more holistic view into their own categories and products combined with the retailer’s understanding of its own business and marketing activities should result in a more accurate overall plan. Best-in-class planning systems can support this kind of collaboration by providing a platform for inputting multiple different forecasts, with the additional capabilities to alert on any differences, edit plans, and finally, disaggregate the agreed plan into the necessary level of detail—stores, products and days—to support operational execution.

7. Conclusion: Team up with the Machines to Win

Retail is in turmoil, and it is unclear what the impact of the different sales and delivery channels, store formats or even retailer players will be. In 10 – 15 years, we will probably look back at this time in amazement and wonder “How did we not see this coming?”

Some predictions about the future of food retail are, however, easy to make:

  1. 1. Food retailers will waste less. It is a disgrace to spend so many resources growing, transporting and handling food products just to have them end up in the dumpster behind a supermarket. Grocery retailers must and will take responsibility for significant reductions in food waste, and, because waste eats profit, their efforts will also be great for their businesses.
  2. 2. The food retail supply chain will become more effective. Consumers have become very price conscious and will not accept premium prices to keep inefficient supply chains in business. No one benefits from wildly fluctuating workloads or capacity requirements caused by poor planning and management, so neither retail employees nor management should be sad to see old, inefficient practices go.
  3. 3. Technology and automation will play a big part in the transformation of retail. We have already seen this in other sectors that once relied heavily on manual routine work. There is no reason why retail would not go down the same path.

To summarize, retail supply chains need to become more responsive and finely controlled than ever before to meet the demand for fresh products with minimum waste. At the same time, retail supply chains need to become more efficient by optimizing inventory flows from multiple perspectives — store operations, distribution, picking and warehousing — to meet the price pressure. This is only possible by teaming up with the intelligent machines.

The world of food retail is too complex to be managed with notepads and intuition. This has, of course, been true for a long time. The breaking news is that not only are the simplest jobs being automated, but significantly more advanced planner roles are being filled by machines, too. More importantly, intelligent automation will not only replace manual work, but take planning to a level of granularity never before seen.

Will there then be any role for humans in this brave new world? Yes, there will be plenty. Three important roles are:

  1. 1. Master of the machines: We are making great progress in specialized AI, the kind of machine intelligence useful for solving very specific tasks. However, we still need to have talented people designing the systems and determining when and how the available machine intelligence is best used.
  2. 2. Colleague to the machines: Machine learning algorithms are very dependent on access to data. They have a hard time applying common sense or coming up with innovative solutions in new scenarios with insufficient data. This is where their human colleagues can provide invaluable insight.
  3. 3. Innovators, thinking beyond the machines: Especially in businesses going through creative destruction, there is a great need for novel thinking, new business models and new ways of delivering food and services to consumers. Retail innovation is still far beyond the capabilities of AI.

So please, do not hold your breath waiting for AI to revitalize your retail business or even solve your supply chain challenges. But please do start phasing in the use of machine intelligence where most feasible and impactful. This collection of best practices is a good place to start.

Authors of this guide

Johanna Småros

Johanna Småros

Group Co-founder, Ph.D (Tech.)
johanna.smaros@relexsolutions.com

Want to know when similar resources are published?

Subscribe to receive a monthly digest of our most valuable resources like blog posts, whitepapers and guides.

Your data is stored only for business-to-business communication purposes. See our privacy policy.