Dynamic Pricing in Grocery Chains (Hypermarkets, Supermarkets, Discounters)

Pricing is a powerful lever for increasing customer traffic and sales.

The most popular pricing tool in grocery retail is promotions and discounts. Today, promotions are an absolute necessity for stores in many countries. Without promotions, customers will stop visiting the store. However, simply relying on promotions is no longer sufficient for stores to expand and succeed in today's market.

Meanwhile, your store has the potential to increase gross profit.
With a 100% probability:

Some of your SKUs are overpriced. Consumers are not willing to buy them at that price because they can find better deals elsewhere (from competitors, larger chains with greater purchasing power, or on marketplaces). By reducing these prices to optimal levels, your store will experience an increase in sales. "Unpopular", previously overpriced products will now generate profit.

Some of your SKUs are underpriced; that is, customers would willingly pay more and purchase the same volume. By raising such prices to the optimal level, your store will sell the same number of units but earn more on each sale.

To unlock this profit potential, pricing automation is necessary.

In this article, we have compiled the most effective tools and practices for automated pricing. These tools and practices have enabled grocery chains to achieve consistent sales growth and gross profit in 2022-2023. We will explain how it works with Imprice in the stores of our clients. After studying this information, you can add similar tools to your pricing software or leverage Imprice's ready-to-use pricing modules.

Reading time: 23 minutes.
#

Increase in Gross Profit: Benchmarks for Retailers Launching Dynamic Pricing

One of the most pressing questions on the minds of retailers implementing dynamic pricing is the extent to which it will drive an upswing in their store's gross profit.
Here are the performance indicators that grocery retailers can expect to witness during the pilot phase:
Profit lift
Grocery,
online
Profit lift
Grocery,
brick-and-mortar
Imprice customer statistics: initial growth during pilots of dynamic pricing
In the post-pilot phase, gross profit continues to increase (current record stands at an impressive +15%).
"Brich-and-mortar" — discounters, supermarkets, and hypermarkets, in-store purchases.
"Online" — Home delivery orders through the store's delivery service.
Factors influencing the magnitude and speed of gross profit growth are detailed HERE.
Hypermarket chain sees
a steady increase in gross profit
and lures customers back from competitors
Hypermarkets increase
gross profit by 6,6%
with AI and KVI pricing
#

Key Characteristics of Grocery Retail for Pricing Automation:

a - Wide product assortment. Hypermarkets typically have around 20,000 SKUs, while supermarkets have around 5,000 SKUs. For non-top-selling items, many prices are suboptimal (either inflated or deflated), because retailers utilize solely fixed markups and promotions. Calculating the optimal price without robust automation is excessively labor-intensive.
Relevant: AI-driven pricing for Foreground and Long Tail segments.

b - In many countries, retailers do not implement electronic price tags. Paper price tags are manually changed, with stores typically replacing 100 to 500 price tags per day.
Relevant: Intelligent queue for price tag replacements.

c - In certain countries, sales of products with promotional prices account for 50% or more of total revenue. Sales of some items are heavily dependent on promotions, that is customers purchasing them exclusively during discount periods. For such items:
-- Without promotional discounts, only occasional sales occur at regular prices.
-- Low regular prices are ineffective.
Relevant: Pricing for promotion-dependent items.

d - Vertical and horizontal product lines exist, including different package sizes and variations in important product characteristics such as taste and fat content. Incorrect or illogical price ratios within a product line can lead to customer aversion or push them towards purchasing smaller package sizes.
Relevant: Pricing for product lines.

e -
Private label brands (PLBs) are present. One chain can own 2-3 distinct private brands that serve different purposes.
Relevant: Pricing for PLBs.

f -
Key Value Items (KVIs) are about 10-15% of the assortment.
KVIs are products that shape customer perceptions of the store's prices and availability of goods ("they have good deal prices and all the products I need" or "everything is expensive there, and they never have the products I need").
Relevant: Pricing for Soft and Hard KVIs. Revealing the optimal price position for Soft KVIs.

g - In recent years, items rotation has significantly accelerated, supply chains are being restructured, new products and brands are constantly appearing on the shelves.
Relevant: Pricing for new products.

h - Customers easily switch to competitors if they offer lower prices.
Relevant: Personalized pricing — a tool for successful competition and protection against cherry-pickers.
#

Key Pricing Practices and Price Optimization Steps — Drivers of Profit and Sales in Grocery Retail in 2023:

1

Stage 1: Analyzing the Assortment.

90% of the commercial success of pricing depends on how accurately you segment your assortment into groups with different "price behaviors" and how accurately you choose a pricing strategy for each group.

It is required to know:
Which SKUs can and should have their prices increased.
Which items require utilizing competitive pricing.
Prices of which products should be calculated together, within a single price vector, to avoid sales cannibalization.
What is more profitable for a specific product role, maximizing gross profit, or revenue, or unit sales.

Thus, it is required to reveal the role of each product in your sales, its impact on the sales of other products, and its price sensitivity to competitors' offers.

In your assortment should be identified such segments of items as:
Customer traffic drivers.
SKUs that drive the sales of other products.
Products that affect the pricing and assortment perception of the store.
Profit drivers.
Substitutes.
Complementary goods.
And others.
Imprice's AI algorithms automatically identify the role of each product, providing retailers with a complete list of your Soft KVIs and Hard KVIs.

Learn more about assortment clustering (revealing product roles) here: Assortment Clustering with Artificial Intelligence
2

Stage 2: Increasing Profit:
Configuring and Implementing Pricing Strategies for Each Product Role and Groups with Special Characteristics.

Competitive Pricing for KVIs.
KVIs are products that shape the customers' perception of the assortment and prices of the store:

If certain KVIs are frequently unavailable on the shelves, customers classify the store as one with a poor assortment, constantly lacking what the customers need.

If the prices of certain KVIs are higher than the optimal level compared to competitors, customers think, "This is an expensive store, it's not advantageous to shop here."


Errors in managing KVIs lead to an increasing number of customers switching to competitors' stores where more KVIs are available, and they are offered at more attractive prices.
If the retailer makes "KVI mistakes", they experience a decrease in sales and profit.
All types of KVIs — Hard, Soft, "false," and "hidden" — are thoroughly discussed in detail here:
KVI In Retail — 4 Common Errors In 2022-2023
To increase the number of customer visits (and regain lost customers), a store must:

a– Know the complete list of its KVIs.
b – Maintain their on-shelf availability.
c – Aware of the competitors' prices for these KVIs.
d – Maintain KVI prices at an optimal position relative to competitors.

How retailers perform the actions described above with Imprice:

a – Know the complete list of KVIs.
With Imprice, you obtain it in Stage 1.
Commonly, KVIs are 10-15% of the assortment.

b – Maintain the availability of KVIs in the store.
The chain and store team's task is to adjust order management according to the complete list of KVIs to ensure a constant availability of the entire list on the shelf.

c - Aware of the competitors' prices for KVIs.
The nature of food retail (discounters, supermarkets, hypermarkets) typically necessitates offline price monitoring (collecting prices from competitors' brick-and-mortar stores). Often, hyperlocal monitoring is required, which involves collecting prices in competing stores located nearby your store (within walking distance or in the same area).

Recommended frequency of KVI monitoring: at least once a week.
The collected prices should be integrated into the pricing system with minimal delay.
Imprice clients utilize specialized software for monitoring competitors' prices. It is a mobile application installed on the "price collector's" phone.
In the competitor's store, an employee opens the mobile app and takes a photo of the relevant price tag. The application automatically recognizes the product and its price from the photo, matches it with the corresponding SKU of the Imprice client, and automatically loads it to the retailer's data management system.
Competitors' prices appear in the Imprice client's system right during the price collector's visit to the competitor's store.

We would strongly recommend using a similar application for offline price gathering.

Our clients' experience shows that if the store team collects competitors' prices manually (the collector writes down prices in a notebook or a file on their phone, then an operator enters these data into the system), the quality of competitive pricing significantly declines. Reasons for this include:
Errors in manual recording and digitization (misreading or mistyping numbers, mixing up products).
Prices appear in the system with a delay of several days. By that time, competitors' prices may have changed.
For discounters and soft-discounters formats, we would also recommend conducting price monitoring for entire assortment once every 2–3 months.
You will be able to see how much your price index deviates from competitors on average and whether it aligns with your chosen pricing strategy.
d – Maintaining KVI prices at an optimal position relative to competitors.

For Hard KVI and Soft KVI, automated competitive pricing is launching based on highly flexible rules:

Imprice automatically uploads offline and online prices of your competitors.
For each KVI, a target price position is set relative to competitors:
first market price, second, third, fourth, fifth.
The target position is determined based on clustering results or according to your pricing strategy.
The Imprice system automatically recalculates prices for your KVIs, considering changes in competitor prices, all internal factors, and constraints (including the minimum allowable markup).

Result:
Prices of your KVIs automatically follow competitors.
Customer traffic and sales increase.
At the same time, you are protected against trading at a loss through markup restrictions.


Personalized Pricing: A Tool for Successful Competition and Protection against Cherry-Pickers.

This solution is suitable for retail chains with an advanced loyalty system, meaning those that have the ability to offer customers a special price tied to specific loyalty cards at the checkouts.
Personalized pricing is relevant for KVIs that are popular among cherry-pickers.
Cherry-pickers are a category of consumers who tend to only purchase products with the best deal prices in each store. By attracting such customers, the store incurs losses.
Personalized pricing is particularly beneficial for small chains.
Example.
A small local grocery chain, Retailer X, sold branded cat food. It was a Hard KVI, that is the product sold well only at the first or second minimum market price:

Price of €1.9-2: 20 units sold per day.
Price of €2.8-2.9: 3 units sold per day.

Large, countrywide operating chains in that city sold that cat food for €1.89.
The cost price of the cat food ranged between €1.8 and €1.9.
Since the cat food was a Hard KVI, some customers were forced to shift to the stores of large chains. This reduced the sales of Retailer X.

To retain these customers, Retailer X attempted to sell the cat food at cost price, €18.9. This attracted a lot of cherry-pickers who were buying only the cat food or the cat food plus other items at a discounted price. The result was a decrease in the retailer's profit.

In addition, with a low price, the retailer could no longer make promotions for the cat food, since they sell the cat food at the cost price, and couldn't offer an additional discount.

Solution.
Analyzing loyalty card data, Retailer X identified a segment of "loyal cat owners" who:
had purchased cat food 3 or more times,
had other items in their receipts.

For this segment, the retailer set for the cat food a special price, at €1.89.
At the shelves, the price was displayed as €2.89.
The price of €1.89 rubles was activated only at the checkout when reading a loyalty card belonging to the "loyal cat owners" segment.

The stores informed their "loyal cat owners" through its channels about the new personalized price. The "loyal cat owners" were satisfied and shifted back from shopping in large chains' stores. Sales and profit of the retailer X increased.

Cherry-pickers could no longer purchase the cat food at cost price.
During promotions funded by the manufacturer brand, the store provided a deep discount on the cat food, lowering the shelf price from €2.89 to €1.89.

AI-driven Pricing for Foreground and Long Tail segments.

For Long Tail and Foreground SKUs, artificial intelligence algorithms optimize prices automatically.

1 – You choose the optimization criterion:

Maximum gross profit.
Maximum revenue.
or Maximum sales in units.
The criterion of "maximum profit" means that the algorithms' task is to find the optimal set of prices that ensures maximum gross profit under current conditions.
2 – The algorithms study the demand dependence for each product on its price:

Analyze the sales history at different price levels.
If the history of price changes is absent or insufficient, the algorithms conduct a series of pricing experiments:
  • Your experts, in collaboration with Imprice experts, set the under and upper limits for the price.
  • Algorithms set prices within the specified range and measure demand:
    1. Set the most likely "optimal" price for the current day or week.
    2. Study the demand.
    3. If the demand increases, try raising the price. If the demand decreases, try lowering it.
    This process continues daily or over other specified periods until the most optimal price is established for a certain period. If the market situation changes, the algorithm immediately reacts at step 2.
  • At the same time, they resolve the "Exploration-exploitation dilemma":
    The task is to spend as little time as possible finding the optimal price and to exploit this price for as long as possible.
  • For substitute goods, the algorithms optimize the entire price vector of the substitute group, not just the price of a single product.
Example of managing substitutes.
We offer 10 types of laundry conditioners that are substitute goods for each other.
On average, the group of conditioners sells 100 bottles per day.
When the price of one conditioner is lowered, the majority of customers switch to it, resulting in a decrease in sales of other conditioners. However, the average daily sales of conditioners remain at 100 bottles.

"Optimization of the entire price vector" means that the algorithms search for a set of prices that increase both sales and profit from the sales of the entire group of substitute conditioners.
3 – Results:

For overpriced products, prices are reduced to their optimal levels. This leads to a significant increase in sales and profit.

For underpriced products, prices are increased to their optimal levels. Sales levels remain the same while profit increases due to improved sales margin.

Pricing of Promotion-Dependent Products.

Promotion-dependent products are mostly sold during promotional discounts, and during regular price periods, only occasional and isolated sales occur.

In some countries, examples of promotion-dependent products are coffee, tea, and household chemicals.
Often, the following holds true for such products:
The volume of promotional sales is influenced not only by the promotional price in monetary terms but also by its difference compared to the regular price.

Example:
A retailer compared sales at the same promotional price but with different regular prices.
1 - Regular price of €1.8, promotional price of €1.2.
2 - Regular price of €1.4, promotional price of €1.2.
In the first case, promotional sales were significantly higher: a deep discount provided better motivation to purchase the product.
To increase profit from the sale of promotion-dependent products, it is required to:
a - identify all such products,
b - set a lower limit for the regular price that ensures a sufficiently deep discount (e.g., 30%),
c - calculate the maximum price for each promotion-dependent product
that is higher or equal to the "lower limit of the regular price,"
that ensures the same sales volume during regular price periods as with the "lower limit of the regular price."

The Imprice platform enables automatic calculation of optimal prices for promotion-dependent products through the following steps:

1 - The analytical module identifies the complete list of promotion-dependent SKUs.

2 - For the products in this list, AI-driven algorithms optimize prices; that is, they calculate prices that maximize gross profit during regular price periods:
the retailer's team set the range within which the algorithms can change prices,
they set the lower limit equal to the lower limit of the regular price, which ensures a possibility to make a deep discount during promotions.

How AI-driven price optimization works is described above.

Pricing for product lines.

In the grocery retail, there are vertical and horizontal product lines (packages of different volumes, variations in important product characteristics such as taste, fat content, and so on).
An example of a horizontal product line is a specific brand of yogurt with different flavors.

Examples of vertical product lines are yogurt (Coca-Cola, cream, canned goods, shampoos, etc.) in different volumes.
Incorrect or illogical price ratios within a product line can lead to customer aversion or push them towards purchasing smaller-sized packages.
Here are examples of logical price relationships:

A 2-liter bottle of cola should cost more than a 1-liter bottle (at the regular price).
2 one-liter bottles of cola should cost more than a single 2-liter bottle.
Deviation from such principles is undesirable:
a) For instance, if a store sells different flavors of yogurt at different prices, and the customer's favorite flavor is priced higher, it will undoubtedly cause annoyance and foster a negative perception of the store.
b) Customers will realize that the store does not support the principle of "buying a bigger pack is more advantageous," and that all prices need to be checked, compared, and recalculated. This, too, will lead to customer irritation.
c) There will be a decrease in consumption. It is known that people tend to purchase larger packages because the price per liter/kilogram is lower. However, having those big packs at home, customers start drinking or eating a larger volume of the product or beverage. In the case of illogical price relationships within a vertical product line, a customer may begin consuming, for example, 1 liter of cola over 2 days instead of two liters. This will result in a decrease in the store's revenue and profit.

Imprice allows for the complete automation of pricing for both types of product lines and automatically prevents the aforementioned negative situations:

1 - AI-driven price optimization for horizontal product lines.
Artificial intelligence algorithms optimize the prices of the entire product line as a whole, rather than individual SKUs within the line.
In the example of yogurts with different flavors, the optimal price for all flavors will be calculated simultaneously. The prices for the entire product line will be the same (if such a strategy was set at the start of optimization).

2 - Automatic rule-based checks for vertical product lines.
It is only necessary to input the rules once.
Then the platform will automatically verify whether the designated price relationships within the product line are maintained.
An example of a rule check is ensuring that one liter is at least 15% cheaper than one and a half liters.

Pricing of new products.

In recent years, there has been a significant increase in brand rotation within assortments. As a result, the task of calculating for new products the optimal price that maximizes store profit is becoming increasingly common.

How to automate this task with Imprice:
  • Step 1. Initial arrival of a new product.
    The category manager sets the price. The store sells the new product at this price for 2–4 weeks.
  • Step 2. Revealing the product's role in store sales.
    Artificial intelligence algorithms analyze the sales data of the product during the initial fixed-price period, and in the first approximation reveal the product's role.
  • Step 3. Launching automatic pricing according to the identified product role.
    If the product behaves as a Key Value Item (KVI) and it is possible to gather prices at which competitors sell it, KVI competitive pricing is employed for the product. Imprice will automatically maintain the price of this SKU at a set position among competitors, such as the second-lowest market price. All necessary constraints will be automatically considered, primarily to ensure that the markup does not fall below the defined level in euros or percentage.
    If the product was identified as a KVI, but gathering competitor prices is not possible, automatic price optimization is launched based on the criterion of "maximizing sales in pieces" or "maximizing revenue."
    If the product behaves as a Foreground or Long Tail, including cases where there were zero sales of the product during the initial weeks, the price of the new product will be managed by AI-driven optimization algorithms. The algorithms will study the demand at different price levels within set price range and determine the price that maximizes profit.

Pricing for Private Labels.

Almost every grocery retail chain has 1-4 private labels (PLs).
Typical roles of PLs are as follows:

a - Economy-tier PLs.
These are extremely low-priced products, usually achieved by significant reductions in ingredient quality.
Mission of such PLs: Attract and retain low-income and budget-conscious shoppers.

b - PLs with high margins, at a price a bit lower than popular brands.
These PLs provide quality comparable to popular trademarks, but at slightly lower prices.
Mission: By enticing customers by slightly lower price, shift them from well-known brands to private label SKUs. Selling private label items is more profitable for the store, as the margin is higher. Thus, the retailer experiences an increase in profit.

c - PLs with unique qualities.
These are "anchor products" favored by shoppers, often serving as the reason for visiting the store. Such products have special characteristics or unique flavors that cannot be found in other stores.
The price of such PLs may even be slightly higher than that of well-known branded products.
Mission: Foster customer loyalty to the store by providing a unique consumer experience.

Imprice offers a set of unique tools that address all private label pricing challenges:
  • Competitive pricing.
    For instance, a store can set the prices of its economy-tier PL in line with competitors' economy-tier PLs.
  • "Original vs. Analog" comparison pricing rules.
    Algorithms automatically ensure that the price of private label item differs from the branded product price by a specified percentage or value in euros.
  • Price optimization for unique "anchor" PL SKUs with AI-driven algorithms.
    The algorithms run a series of pricing experiments to determine the price at which the store achieves maximum gross profit from the sale of unique PLs.
3

Additional Pricing Practices and Tools
for Profit and Sales Growth.

Refining optimal price positions for Soft KVIs and competitive products.

The AI-powered module can help further increase sales of Soft KVIs and relatively low-selling items that customers tend to compare to competitors' offerings.

LEARN MORE

Price tag replacement: Intelligent queue.

If a store doesn't utilize electronic price tags, how does dynamic pricing work, and how are price tags changed? Commonly, in grocery retail stores with paper price tags, there's a limitation: it's possible to change 200-500 price tags per day.

Imprice has the "Intelligent Queue" tool.
The algorithms automatically determine which price tags are crucial to change as soon as possible and which can wait.
At the top of the list are scarce, well-selling products that have a significant impact on gross profit.

Examples of indicators that determine the "urgency of price tag replacement" in the algorithm:
Procurement conditions for the product were changed, and markup significantly decreased or even became negative.
The product contributes significantly to the store's total revenue.
According to forecasts, the product will strongly affect the store's profit.

LEARN MORE

Automated "Low Price" Price Tags List.

Typically, stores print price tags with "best deal prices" on colored paper, either red or yellow, to capture the attention of shoppers.

Imprice can automatically generate and send to the store two separate lists of new price tags for printing:
1 - Best Deal Prices. For printing on colored paper.
2 - Regular Prices. For printing on regular paper.

This saves the store employees' time and eliminates errors such as printing favorable offers on regular paper.

On-shelf Availability Control — Ensuring Product Replenishment on the Shelf

"Product availability in the store's warehouse" does not equate to "product being displayed on the store shelf."
Example:

A supermarket chain ran a dynamic pricing pilot. The team monitored an extensive list of store performance indicators on a daily basis.
The pricing analyst noticed recurring situations: a good selling product suddenly could stop selling, despite it being in stock. They created a complete list of such SKUs, and studied it with store employees. In nearly 100% of cases, it was discovered that the supplied goods were left on pallets in the warehouse. It took a few days to place such goods on the shelves in the trading hall. Since the products did not reach the shelves, it resulted in zero sales for the SKUs for several days.
To address product unavailability on the shelf, Imprice offers a simple yet effective tool, an automatically generated report "Abnormally Paused Sales."

The retail chain receives the following information from the report:
Heineken's beer, average daily sales: 3.85 cans, in stock: 586 bottles, were minimum three days in stock, zero bottles of Heineken's sold in the past two days, three units sold within a week. On-shelf availability issues are suspected.
Cat food, average daily sales: four units, in stock: 185 units, were minimum three days in stock, zero units sold in the past two days, two units sold within a week. On-shelf availability issues are suspected.
And so on.

Despite its simplicity, this report accurately identifies the problem and enables effective resolution. The store responsible employee manually checks the list of SKUs from this report on a daily basis to assess the shelf situation.
Practice shows that with regular active monitoring of the items in the report, the number of "signals of paused sales" significantly decreases.
Talk to Imprice pricing experts: