Discover How Imprice Can Help You Achieve Your Business Goals
We will make a detailed demonstration of Imprice and use cases:
Discover How Imprice Can Help You Achieve Your Business Goals
We will make a detailed demonstration of Imprice and use cases:
AI-Driven Price Optimization
Rule-based dynamic pricing is a quick way to increase gross margin and raise sales.

If you need additional growth of business results, the next step is ai-driven price optimization based on demand sensing:
Launch the Imprice "Automatic KVIs Identification" module
One algorithm of the module identifies all KVIs in your assortment and automatically updates the resulting list in real-time.

Another algorithm calculates the optimal competitive position for each KVI. The result is a transparent, data-based recommendation for every specific KVI: what is more profitable for your company, to use the first minimum price of the market, or second, or even fifth.
For each product, get a list of your key competitors that affect your sales with the Imprice "Key Competitors Automatic Identification" module
For each item, the module will point out:

— A list of key competitors: retail companies that affect your sales of the item. The analytics use the history of competitors' price changes, your sales history, other factors.

— A list of competitors you need to exclude from your price calculations because they don't affect your sales.
Оptimize your prices automatically at category-level or assortment-level, and maximize your sales or gross margin
The module "Automatic price optimization, maximizing a selected criterion" with the help of real-time demand sensing calculates the optimal sets of prices with those you get the maximum gross profit from your assortment or specific category under current market conditions.

If your business goal is your market share increasing, you set another optimization criterion, maximizing revenue or sales in pieces.
The solution also allows you to find optimal prices for new items rapidly.
Module 1. Demand sensing-based price optimization, maximizing a selected criterion: gross margin, revenue, or sales volume in pieces.
Method details:
The module optimizes prices for both offline and online stores.

Studying consumers' reactions and leveraging vast sets of real-time data, machine learning algorithms improve prices fast step by step. For example, for online stores, algorithms also analyze a website and mobile app visitors' behavior.
Based on demand sensing, the module adjusts SKUs' prices to maximize sales or maximize gross margin — depending on what goal you have set for the platform.
In particular, for online stores, it considers item's page views, bounce rates, conversions to order from an item page, average conversions by SKU, brand, category.

Generally, it is crucial to optimize prices at the entire category-level or assortment-level; it helps avoid sales cannibalization within categories and considers SKUs' cross elasticity.
However, if necessary, you can use the machine learning module to find the optimal price of a specific SKU.
The advantage of price optimization with Imprice:
It considers any digitized business constraints affecting your sales or a combination of such constraints.

Constraints examples:
limits of stocks,
fixed advertising budget,
limited delivery capacity.

The algorithms calculate a set of prices that maximizes your profits under the given constraints at assortment-level, category-level, or SKU-level.
Main use cases:
Machine learning algorithms analyze:
real-time and historical sales data,
for e-commerce, data of website and app visitors' behavior,
how consumers react to goods and their prices,
other factors,
and rapidly find the set of optimal prices with which your company gains the maximum profits at assortment-level, category-level, or specific SKU-level, under current conditions.
ML-based pricing methods are efficient to:

1. Optimize prices of Long Tail, and even the entire assortment, except for KVIs.
For KVIs, we would advise using KVI Pricing.

2. Find the optimal price for new items that don't have any sales history in your assortment.
ML algorithms rapidly explore and calculate the optimal price that will maximize sales or profits.

3. Find the optimal price for items for that you don't have competitive environment data because ones don't have equivalents or other reasons.
The business goal of your company — to increase market share rapidly or maximize gross margin — determines what price optimization criterion you have to choose.
What do retailers achieve with AI-driven price optimization:
Price is one of the most important factors for retailers.
For 44 to 66% of consumers, price is a very or extremely important factor in determining where to shop, depending on the retail segment.
(source: consumers and retailers research, 2021

When prices are higher than what consumers are willing to pay, shoppers leave the store to buy elsewhere. If prices are too low, the retail company loses its profit. Even worse, shoppers can perceive a low price as suspicious. They think that the low-price product is a fake or poor quality good, so they don't buy, and your sales and gross margin decrease.

The optimal price is the price that gives you the maximum gross margin if your company needs to increase profits.
And if the company needs to raise market share, the optimal price provides maximum sales, maintaining an acceptable margin level.
Generally, three factors determine the retail price:
- cost price,
- competitors' prices,
- value that consumers gain with a purchase in this particular store.

Shoppers can perceive service factors as a value.
Service factors examples:
fast delivery, convenient payment, user-friendly website: good search engine, excellent photos, helpful descriptions, shoppers reviews about each SKU and its features.

Psychological factors also can produce additional value: shoppers' confidence in store's reputation, image, positive experience, fear of fakes in other stores, self-identification - "I buy here because I belong to a certain social group".

Consumers are willing to pay more for such value.

But how much more?
Commonly company managers have to use their intuition to answer.
Machine learning algorithms allow identifying exactly how much customers are willing to pay your store and with what price your company reaches its business goals.
AI-power module identify:

what prices and how much your store can raise to maximize gross margin,
what prices your store has to lower for impactful sales growth or reigniting demand.
ML algorithms identify "suspicious" probably overpriced items among SKUs with poor sales and manage a smooth step-by-step price reduction for ones until conversion rates improve.
As a result:

1. The item's price lowers slightly, so sales of the "suspicious" SKU raise significantly, improving the company's inventory turnover.

2. Or the "suspicious" item has no potential for good sales, even at the minimum margin level, because of poor quality, unattractiveness, bad reviews.
The platform marks it as illiquid and removes it automatically from the assortment matrix with a ban on purchase. That prevents the accumulation of illiquid stocks.
If your target is market share increasing, the platform calculates the optimal set of prices for that, considering business constraints and maintaining margin as possible.
Suppose our margin constraints let us set the second minimum price in the market. But the level of sales is almost the same at the second, third and fourth minimum prices. Only if we set the fifth minimum price, sales begin to decrease.

Under these conditions, the algorithm will set our price at the fourth minimum market position.
Case study:
August 17, 2020 is a Russian retail chain and online store of sports nutrition. The company's consumers mostly are active visitors of Moscow or Saint Petersburg gyms, improving the training result with sports nutritional supplements and vitamins.

Due to the COVID-19 spring lockdown, gyms in Moscow and St. Petersburg were closed until July 2020. As a result, the company's offline stores were closed, and with non-worked gyms, online sales decreased almost to zero. AI-driven Imprice modules helped to reignite demand and accelerate the gross profit of the online store.
Module 2. Automatic KVIs Identification. The optimal price position for KVIs comparing to competitors
1. One algorithm of the module identifies all KVIs in your assortment and automatically updates the resulting list in real-time: adds to the list new KVIs, removes items that have become non-KVIs.
Common mistake:

There is a simplified way, which identifies KVIs as top-selling items in a category, or as SKUs bring the most significant part of revenue.

This approach is a bit dangerous.

First, not every "top" is a KVI.
For example, your top-selling item is a product that competitors don't have in their assortment. It is not a KVI; its optimal price only depends on cost price and demand.

Second, probably your "not top seller" is a latent KVI.
For example, item X has appeared in your assortment several times in small quantities. It sold out quickly every time. Due to small sales volumes, you have not identified it as a top-seller of the category, although it is a KVI.
The Imprice module identifies all KVIs in your assortment.
For each item in your assortment, Imprice algorithms analyze a broad set of factors and identify if it is KVI or not. Examples of such factors:

Sales volume of the item.
Inventory turnover.
Competitors' assortment.
How the item's price changing in your and competitors' stores affected sales of the item.
Quantity of people who had an interest in the item on the website.
How often visitors leave the website after the item page viewing.
How price affects conversion to order.
Whether consumers buy other products with the item.

The result of algorithms:
Complete analysis of all SKUs.
A mark in a particular item field: is it KVI or not.
A segment that includes all items marked "KVI".
Setting up special competitive pricing rules for that KVI segment in Imprice takes only a few minutes.
2. Another module algorithm calculates the optimal competitive price position for each KVI.
The result is a transparent, data-based recommendation for every specific KVI: what is more profitable for your company, to use the first minimum price of the market, or second, or even fifth.
KVIs in the assortment of the same store may have a different price elasticity of demand.

A frequently used pricing strategy is to keep all KVIs at the second-lowest price position on the market or among key competitors.
This strategy is usually efficient: it generates good sales and is not as dangerous as the first lowest price.
First Lowest Price:
can result in a price war,
cuts margin, often unnecessarily, when with the second price position it's possible to sell the same quantity in pieces,
attracts many unloyal shoppers who make purchases only once and buy only KVIs due to the low price.

Though the strategy "the second minimum price for all KVIs" has many advantages, there is a more effective strategy: an individual calculation of the optimal price position for each KVI.

For example, for some KVIs, consumers can have extremely high sensitivity to price, so switching from the second minimum price to the first can lead to a surge increase in sales.

Contrary, the competition level may be lower for other products, and shoppers may appreciate some unique advantages of the store precisely for buying this type of product.
For example, an online store is one of the few that provides the option of fitting before payment. In such cases, with the second, third, fourth, and even the fifth minimum price, the store would gain approximately the same number of sales in pieces. So with the second price position, the company loses significant margin and gross margin.
For each KVI, the Imprice algorithm:
analyzes historical data: sales, stocks, competitors' prices for previous months,
compares the KVI's price with competitors,
checks the sales volumes with every price position,
calculates the optimal price position of this KVI comparing to competitors.

Optimization criterion recommended: maximum revenue.
The algorithm considers both KVI sales and sales of goods that consumers bought with that KVI.
When the KVI is a strong basket driver, and in some other cases, another optimization criterion may be more efficient, maximum sales in pieces.
3. You will set up pricing rules for KVIs in a few clicks, and then KVIs price recalculations will work automatically.
A hint:
Combine KVI pricing with performance marketing optimization. Load KVIs prices automatically directly into your ads from Imprice.

Module 3. Automatic identification of key competitors.
Key competitors are stores whose pricing affects your sales significantly.
Why is identifying key competitors crucial for KVIs?

Suppose you consider the prices of all stores in your location for your KVI's pricing.
Two stores sell this KVI at a price much lower than other retailers.
These stores don't affect your sales:
One store offers delivery time that is unacceptably long for your consumers: they want to receive the product quickly.
Another store has horrible reviews. They process orders for a long time; they can bring poor-quality or fake goods. Your consumers prefer to don't have deal with such a store.
If you consider these stores' prices in your pricing and set your price close to theirs, you lose margin and gross profit unnecessarily.

Another example.
You are a large retailer with a broad assortment. You monitor the prices of other large retailers and consider only their prices in calculations.
A small online store sells only a narrow set of KVIs at attractive prices and actively uses social media blogs for promotion. They have qualified staff, fast delivery, great reviews and more aggressive prices than yours. More and more of your customers choose to buy these KVIs in that small online store, getting more attractive prices and a positive customer experience. Your KVIs' sales are reducing. Sales of goods with a high margin, which shoppers bought with these KVIs, are also decreasing.
Consumer price awareness is extremely high nowadays; the number of KVIs in a category can reach 30-50%. The list of key competitors may differ for each KVI. These are consequences of digitalization.

It is impossible to analyze the complete list of competitors for each item manually or in a semi-automatic mode and calculate for every competitor does he affect your sales or not. Non-automatic calculations push to simplifications:
to monitor a narrow list of the most visible competitors,
or to "carpet bombing": setting the most aggressive prices, considering all market retailers.
The first causes loss of sales. The second leads to an unnecessary loss of margin.
Artificial intelligence helps solve this problem.
Using machine learning, algorithms:

1. Analyze the accumulated internal and external sales data, stocks, demand fluctuations, internal and external price changes.

2. For each item, build a list of competitors that affect your sales significantly.

3. Update these lists automatically when factors change.
Case study:
August 17, 2020
+23.0% Revenue growth
due to automatic identification of key competitors