Supermarket chain turns to data-driven pricing to add value to consumers and increase profits

CUSTOMER
PROFILE:

Industry:
● Grocery

Location:
● Eastern Europe.

● 3 supermarket chains

● 115 brick-and-mortar stores

● 45,000 SKUs

Founded:
● 2015.

In August 2021, the company launched a pilot test of dynamic pricing in one test supermarket, comparing its business results with a control supermarket.

Key results

The results for August-November 2021 were as follows:

● The company gained an actual dynamic list of KVIs. The list contained 4 times more items than the initial "pre-pilot" list.

● +5% Gross profit growth (the test supermarket's result compared to the control supermarket).

● According to customer surveys, the score of customer satisfaction increased significantly for the test supermarket.

● The company turned its pricing into a transparent, agile and scalable process, easily managed by only one pricing analyst.

Manual pricing didn't ensure
customer satisfaction

CHALLENGES:

● Launch data-driven pricing

● Efficiently identify the complete list of KVIs

● Recalculate KVIs' prices automatically, considering
competitors' prices

● Provide optimal prices to enhance customers' price perception

Since 2015, the company has provided its customers with affordable, everyday low prices on groceries and goods.

In 2021, the retailer's team aimed to adopt a more data-driven approach to its pricing. It was caused by the following facts:

● Competitive pressure increased.

● The company's existing Excel-based pricing tools were not always sufficient to serve customers' needs; sometimes surveys indicated customers were not satisfied about the retailer's prices.

● Pricing assistants manually gathered prices from
competitors' stores and manually entered them into
spreadsheets. The process was long, and there were issues with data quality.

● The retailer did not have a dedicated pricing team. Instead, pricing was handled by category managers, and they were overworked.

● The KVIs were identified using an intuitive and expert
opinion-based approach.
Due to that and data quality issues, it was hard to ensure the targeting price position on the market and to manage price perception.

Searching for the right pricing solution

KEY CRITERIA:

1 - Technical leadership.

2 - AI-driven KVI pricing.

3 - Transparency in both AI-driven price optimization and
rule-based calculations.

4 - Agile zone pricing.

The company's experts compared 5 pricing platforms; it took 4 months. Finally, they chose Imprice, according to the following key criteria:

1 - Technical leadership.
Some platforms did not suit the company's goals because they required manual operations for data exchange. With Imprice, all data exchange processes worked fully automatically.

● Imprice automatically collects competitors' data and or automatically loads competitors' data from retailer's sources.

● The platform automatically loads data from ERP and other sources for every SKU: demand indicators, costs, inventory turnover, stocks, supply dates, and other data.

● Imprice automatically recalculates prices, using rule-based pricing and ai-powered price optimization, and automatically uploads new prices to ERP, data exchange system, website, and other systems.

2 - AI-driven KVI pricing.
It was essential for the retailer to find a solution that would identify KVIs precisely.

Imprice AI-powered analytics module identifies all KVIs in the assortment and ranks them according to the recommended price position against key competitors.

/AI — artificial intelligence/

3 - Transparency in both AI-driven price optimization and rule-based calculations.
When the retailer's experts explored Imprice's features, it was clear the platform provided the required pricing transparency and reporting.

Imprice provides perfect tools making the pricing process transparent: a set of dashboards, charts, reports, plus a complete history of all changes for any segment of goods, any specific SKU, and any pricing rule and pricing logic.

It takes just a few clicks to check:
● what the price was of any SKU in any city on any day,
● how that price was calculated,
● what the supplier's price was on that day,
● what the margin was,
● what prices competitors had.

4 - Agile zone pricing.
Imprice empowered pricing specialists to take each supermarket (or a group of supermarkets) as a different market with a tailored pricing strategy.

With Imprice, each local pricing strategy considers the location's unique characteristics:

● local KVIs and profit drivers list,
● local competitive environment,
● local competitors' pricing behavior,
● local company's target, such as market share raising or profit maximization,
● local Inventory turnover, stocks in local warehouses, costs, supply dates, ...

KVIs dynamic list

SOLUTION:

● AI-powered Analytics

● KVI pricing, competitive pricing

● Competitors' price monitoring automation

● Product line pricing

● AI-driven assortment-level price optimization

After technical integrations were complete, pricing experts ran the Imprice AI-powered analytics algorithms to identify the role of each SKU in the supermarket's assortment:
● KVIs, hard and soft types,
● profit drivers,
● complementary goods,
● substitutes, and others.

The first "AI-made" KVIs list had an approximately similar number of SKUs as the "pre-pilot" expert opinion-based list.
But these SKUs weren't the same:

✓ The "pre-pilot" list missed about 50% of the AI-selected KVIs. So these SKUs had prices higher than optimal for consumers, and the retailer had been losing customer satisfaction.

✓ About 50% of "pre-pilot KVIs" weren't true KVIs. So these SKUs had prices lower than optimal, and the supermarket had been losing its profits.

During the pilot test, the Imprice AI-powered algorithms
reanalyzed the assortment every two weeks (then once a
month), updating the KVIs list: adding "newborn" KVIs and
removing SKUs that stopped performing as KVIs.

Competitors' price monitoring
improving

For each SKU, Imprice began automatically collecting key competitors' prices on the largest domestic e-grocery service which delivered goods from supermarkets.

Pricing assistants manually gathered prices at brick-and-mortar competitors' stores located near the pilot test supermarket; but this process was significantly improved:

1 - Now they collected only KVIs prices.
2 - They use a mobile app. The app showed them a task:
the product for which they had to scan its price tag (to make a pic of that price tag). Then the app automatically recognized the price and sent data to a spreadsheet.
3 - Imprice automatically loaded this collected data into the
pricing platform.
Thus, data collection performed faster, the process costs were reduced and there were no more data issues.

KVI pricing. Product line pricing

For each KVI, the AI-driven algorithms calculated the most profitable price position relative to key competitors. For some KVIs to keep the lowest price was crucial. For others, the second, third, fourth place or average market price was more profitable.

Pricing experts created a special KVI segment and entered into Imprice special competitive rules, that considered set business constraints and the most profitable price position for each KVI.

As well, it was considered in pricing rules if the item was a part of a product line. Consumers could not appreciate if different tastes of the same yogurt would have different prices. Due to that, if one yogurt of the product line was a KVI, rules set prices of all yogurt from that product line equal to KVI yogurt price.

The platform started to calculate prices automatically according to these rules.
Now all price perception drivers had their optimal prices.

AI-driven price optimization

For 83% of tested SKUs, pricing experts launched AI-driven assortment-level price optimization.

How it works in Imprice:

● The module finds the optimal set of prices under current conditions. "Optimal" means, implementing that price configuration, the company achieves the maximum gross profit at the assortment level.

● Machine learning algorithms evaluate goods' cross-impact and segment items by their roles in the consumers' basket: basket drivers, traffic drivers, profit drivers, complementary goods, substitutes, and others.
Then each segment gets its optimal pricing strategy according to your business goals.

● The solution also allows to optimize the prices of a specific category of goods and rapidly find optimal prices for new items.

It was considered in AI-optimization rules if the item was a part of a product line, as well.

Rollout

SCALING:

1 - Across 30 supermarkets of the pilot test chain

2 - Across other chains of the company

In 4 months, the pilot test was complete. The results were as follows:

● The company gained an actual dynamic list of KVIs. The list contained 4 times more items than the initial "pre-pilot" list.

● +5% Gross profit growth (the test supermarket's result compared to the control supermarket).

● According to customer surveys, the score of customer satisfaction increased significantly for the test supermarket.

● The company turned its pricing into a transparent, agile and scalable process, easily managed by only one pricing analyst.



The retailer's team began the rollout:

Rollout step 1. Scaling dynamic pricing across all supermarkets of the chain.

Rollout step 2. Adopting dynamic pricing across the company.