Price is intimately linked to a retailer’s commercial attractiveness – and therefore to its competitiveness, success and sustainability. The pandemic context confirms it: for a retailer, being able to set the right Price at the right time is a question of survival. As a result, the competitive pressure on price is stronger than ever. A differentiating and attractive pricing strategy for the consumer cannot, today, leave the competition behind. This is where effective product matching comes in.
Product matching: competitive comparability as a pillar of pricing strategy
Define your prices according to your competitive perimeter
Collecting competitive data – prices, labels, images, EANs – is the prerequisite for any product matching. For offline and food retail, data is usually acquired from panelists. It can also be collected online, by extracting data from competing websites (Web Data Collect) or at points of sale (Scan In Store). While the raw data collected in this way is of limited use in its current state, product matching can be used to add value to it.
Artificial Intelligence for efficient product matching
Although essential in any competitive pricing strategy, the product matching process is generally tedious. It often requires the user to chain products by product. Depending on the number of products and competitors, the task can quickly become time consuming and discourage your teams.
Automation of EAN-based product matching
Automatic chaining requires no human intervention. The algorithm analyzes the imported competitor data and compares it with the brand’s product data. When it identifies identical EANs, it links the corresponding products together. This chaining is automatic.
Obviously, this process allows our customers to save time and performance on matching. It contributes greatly to generating a reliable and qualitative product database, useful for improving their competitiveness and price positioning.
Depending on the retailer’s sector of activity, it is estimated that between 40 and 80% of its products can be matched. However, automatic matching is not always possible. It is then up to the user to check that the products to be linked are identical or comparable. In this case, the AI supports his decision making.
Manual chaining in one click
Several cases require the intervention of a user, in particular when the products are :
- Identical but their EANs do not match,
- Comparable or similar because they come from the same brand,
- Identical but their size, capacity, volume, etc., do not match.
Here again, Artificial Intelligence supports simplified decision making and accelerates matching. Let’s see how.
The Optimix Pricing Analytics Matching module allows for easy chaining on two axes: product photos and product labels.
The objective: Match products in a single click.
To do this, the AI compares the photographs and labels of the products in the database. It standardizes the names of competing products using a dictionary of synonyms, then assigns them a proximity coefficient weighted by price. Based on this scoring, it sorts the competing products and matches them with the most relevant products of the brand. After that, all the user has to do is determine whether the products are identical, comparable or comparable with a coefficient. With just one click, he can then proceed with the matching.
Artificial Intelligence works on the principle of machine learning. In concrete terms, this means that the more it contributes to the matching process, the more efficient its analysis and scoring will be, further accelerating the matching tasks.
It is possible, upon user request, to define a threshold from which the AI automatically matches products. For example, we can define that the AI will match all products with a proximity score higher than 98%.
If, in spite of everything, the retailer is reluctant to mobilize high value-added resources for the few remaining manual tasks, Optimix provides the necessary resources to subcontract the matching.
Fully customizable reporting
The module has a space dedicated to the statistical analysis of the matching results. For example, you can find data on the level of frontality or the percentage of products matched by the brand. The reporting is fully customizable to adapt to your scope and needs.
