How to improve turnover with effective fresh food inventory management

The fresh food category – fruits and vegetables, as well as meat, seafood, dairy, and baked items – generally account for up to 40% of grocery store earnings. They are also important drivers of client loyalty and store traffic. Most grocery retailers today face a lot of challenges when it comes to fresh food inventory management. This is typically due to the high level of waste involved, which can be due to overstocking, caused by lack of visibility to inventory, as well as out of stocks, leading to missed sales.

In this article, we would like to share the specifics of the fresh category in terms of inventory management, as it’s one of the most sensitive categories and always requires additional attention. 

Our company's team LEAFIO AI Retail Solutions has extensive experience in automating inventory management in retail. We help retailers manage all product categories at all levels of the supply chain: in stores, regional warehouses, and distribution centers. 

With our experience in implementing inventory management automation projects in this category, we have identified several rules and patterns and added them to the algorithms of our system so that our customers achieve economic success in the fresh category. 

The fresh category is a rather broad one and consists of different subcategories. To manage fresh products well, you need to understand not only the general features but also what the differences between these subcategories are. 

• Dairy products

This subcategory is one of the most competitive. Dairy products are everyday goods, so they are traffic-generating products. There is a huge number of large vendors who produce milk, yogurt, and cheese, but local vendors do that as well. This diversity leads to increasing competition between brands. Vendors keep their prices at relatively the same level, and this reduces margins. Besides that, such an effect as cannibalization takes place a lot in this subcategory during promotional activities. Also, dairy products usually have rather a short shelf life - from 3 to 10 days.

• Fruits and vegetables

The most complex subcategory in terms of inventory management among all products in the fresh category. Seasonality plays a huge role here. As a result, vendors have different pricing policies. Purchasing managers compare prices for the same SKU and create the order based on the best pricing conditions. And two batches of goods might be ordered from two different suppliers within the same week.

Related: Multi-echelon Supply Chain: Optimizing Inventory Management

Another negative phenomenon that takes place directly in stores is an inconsistency that happens during the inventory count. Sometimes while weighing the product, customers mistakenly choose a different SKU code, and as a result - wrong balances in the ERP happen. But the biggest pain point in managing this subcategory is the appearance of the product on the shelf and its expiration date. Two products from the same batch may differ, one item may look like the one to be written off, and the other one still looks good.

• Bread and bakery

In some ways, this subcategory is similar to dairy products in terms of management. Bread is also a traffic-generating product, but the difference is in pricing. Bakery products can be divided into two types: own production and social products. The maximum margin for the social group is 5-7%, while the own production bakery margin can reach 50% or more.

One of the main features of this category is the frequency of orders. Bread should be fresh and crispy. Deliveries are done every day or even twice a day. Shelf life is up to 3 days on average.

• Own production, fresh meat, and fish.

Two subcategories with relatively common characteristics might have cannibalization during promotions. If regular tomato sauce meatballs are in the promotion, no one will buy the ones with cheese. The same logic works for meat or fish. If dorado is sold at a discount, demand for sea bass or trout will decrease. Also, sales depend a lot on how accurately and nicely the layout is made.

What are the key characteristics of the fresh food supply chain management?

In general, there are the following special characteristics of the fresh food supply chain management: 

• First of all, fresh products differ from others because of their short shelf life. Often, it is no more than 30 days. Because of this feature, fresh is considered the riskiest category in groceries. Short shelf life requires a very precise balance between availability and write-offs because striving for high availability can significantly increase write-offs, which will immediately hurt the company's financial results. 
• The second peculiarity is a large share in the company turnover. For some retailers, this number can reach up to 60%. And at the same time, the number of SKUs can be smaller compared to other categories. Not having the highest margin, such products generate significant sales volume, bringing the consumer back to the store regularly.
• Third, the category is distinguished by the peculiarities of orders and delivery. For fresh goods, orders are sent at least 1-2 times per week. Mostly, it’s direct deliveries from suppliers to the stores. And the most important thing is the peculiarities of logistics because, during the transportation, special temperature conditions must be kept. 

Recommendations on how to improve the turnover by managing the fresh inventory category

With the proper management of fresh inventory, a retailer can improve the turnover in general, earn more and attract customers to the store since the fresh category is a traffic-generating one. So let's talk about handling this sensitive fresh category and earning more without additional cost for demand stimulation.

1. All about data. Clean and correct data on balances, sales, and write-offs. 

Data accuracy determines the final quality of the order because, with incorrect data at the input, good quality can’t be expected at the output. In terms of fresh goods, there are often problems with the accuracy of balances due to late fixation of deliveries in the ERP, late write-offs, and returns to suppliers (if they are possible according to the agreements and company policy). Also, especially in the category of fresh fruits and vegetables, there is often an inconsistency during the inventory count, so the final balance for the category might be correct, but not for each SKU in particular.

So the main recommendations in terms of improving operational excellence:

• Establish time limitations for fixation of the goods delivery in the ERP. In this case, it is possible to minimize the risk of “negative” balances. 
• Determine the schedule for the inventory count, according to which the count for fresh goods is done more frequently than for dry goods. 
• Establish a time limitation for backdating documents if your company has such a practice. Even the presence of strict rules can stimulate a more responsible attitude to change information from the past periods. This is important because when you change something in the past, it is possible that you will not be able to analyze why certain decisions were made. 

In the Leafio company, we have cases in our project implementations where we are investigating the reasons why the system calculated an order that was mismatched with the manager’s expectations.

2. Consider the remaining shelf life and predict the shelf life of the current stock balance at the time of delivery of the next batch of goods. 

It's no secret that creating and maintaining batch accounting in food retail is impossible since in one shipment may be the goods from different batches with different expiration dates. So it’s very hard to keep track of from what batch the goods were sold and with what expiration date.

But it is important to consider expiration dates, because ignoring this information can cause write-offs, especially for the goods with bad inventory turnover. 

Shelf life should be taken into account at the time of calculating the forecasted balance on the date of the goods delivery. It is important to understand whether the balance will be valid (and appropriate for sale) at the time the next order arrives. 

So our recommendations will be the following: 

• To consider the residual shelf life - the number of days starting from when the product is delivered to the store till the end of its shelf life. This information can be considered to calculate the forecasted stock balance for the product's delivery date. 
• Consider the sales by the LIFO method - last in first out - assuming that the customer will buy the freshest goods first. This approach considers the fact that the entire balance won’t be available for sale even if the goods were not written off in time. Unfortunately, we often face untimely write-offs in implementations and it takes time to change the process. Therefore, we recommend doing a balance check using the LIFO method to minimize the impact of incorrect balances. 

3. Consider variability across the week. 

Every product has its variability across the week. A good example here can be that home appliances are sold significantly better on Friday and weekends than during weekdays. But this pattern characterizes the fresh category as well. And since goods in this category are frequently supplied - daily or several times a week - and the shelf life is limited, we recommend using this week's variability coefficients to predict demand for the next delivery. These coefficients can be calculated by statistical methods or by more advanced methods using AI. 

Considering such coefficients allows a more accurate forecast of the demand. But it should be included in the auto-ordering algorithm because otherwise it might be labor-intensive and can generate a lot of errors. Every product in every store will have specific weekly demand, which depends on its location and format. 

Also, it’s worth mentioning that it is important to consider the time during the day when the delivery is made. This will allow the delivery schedule to be made more accurately and takes into account the required number of days in the forecast. We call this a delivery slot and it helps us make the order even more accurate. 

4. Zero balance for the end of the day for the ultra-fresh category. 

Ultra-fresh goods have some interesting specifics. In a lot of cases, there is no stock balance by the end of the day. In practice, it looks like this: goods are ordered, received, and sold on the same day. But it’s hard to understand if there was enough quantity of the product for each particular day to satisfy the customer demand.

• First of all, for such goods, we should consider the time of day when the last sales were made according to the receipts. This allows you to understand whether there was enough stock balance by the end of the day and whether the demand for the day was covered. If not, the order quantity should be increased to have enough stock balance to cover the demand. But this is only relevant if the availability of goods has not been reduced on purpose to minimize write-offs. 
• OOS is an important indicator for the category, but if the stock balance is zero each day, it becomes more complicated to calculate it. Therefore, another assumption can be made for a correct calculation: if there was not enough balance to calculate lost sales at the end of the day, it is worth using the statistical ADU. 

We would like to emphasize again that for ultra-fresh it’s critically important to have correct data on inventory and sales. 

5. The increasing role of promotion in the fresh category. 

Forecasting and promo management is already a difficult process, and taking into account all the above specifics of fresh, it turns into a very complex, risky and expensive process with a lot of mistakes. 

Therefore, we recommend:

• Use demand forecasting algorithms in promotions that take into account all of the above-mentioned specifics of the fresh category and the cannibalization of demand within the category for the most accurate demand forecasting.
• Take into account actual promotional sales in real-time as the forecast may differ from actual promotional sales. And the sooner you start considering reliable statistics for a particular product in a particular promo, the better. 
• Be very careful with additional promotional layouts at stores. Additional tools are needed to automatically consider that it’s additional equipment to make sure that there won’t be overstock when the promo ends.

6. And last but not least - Ongoing analysis of the bottlenecks in ordering in fresh food order management system 

Let’s focus on the indicators that are specific to the fresh food order management system.

Taking into consideration the mentioned specifics, it’s important to analyze: 

• Writes-off and its dynamics.
• Bottlenecks are caused by the situation when MOQ is higher than sales for the term of the product’s shelf life. 
• Bottlenecks when the number of days for delivery is higher than the residual shelf life.

The Leafio solution has a separate block of reports for analyzing the fresh category. 

For example, this particular report shows specific SKUs in specific locations for which the minimum order quantity is bigger than the average number of sales per shelf life of the item. The system also calculates the estimated percentage of sales of the supplier's packaging and predicts write-offs of that product based on that logic. 

In this report, we can see last week's write-off history, sorted in decreasing order by SKU at storage locations. 

“What gets measured gets managed”

Such an important category as fresh requires precise analysis and control. Key outcomes of it:

• Fresh is one of the most difficult categories in terms of inventory management. And each subcategory requires a separate approach.
• Bad input - bad output. Multiply by 10 and you get the impact of data quality on the quality of inventory management of the fresh category. That's why it's important to do paperwork and inventory count in time and limit document backdating. 
• A specific approach is needed for estimation balances at the projected order arrival date. We recommend doing it by their residual shelf life using the LIFO model. 
• Considering that fresh is a risky group of products, we recommend applying coefficients of variability across the week. 
• For calculating statistics and out-of-stocks for the ultra-fresh category, it’s important to understand the reason for zero balance at the end of the day. Was the balance not enough to cover the demand?
• Advanced tools should be used to forecast promotional demand and should be based on reliable, current statistics during the ongoing promotions. 
• Ongoing analysis and monitoring of KPIs are needed, especially such KPIs as availability, write-offs, and analysis of various bottlenecks that are specific to this category of products.