Product Matching
The Price is closely linked to the commercial attractiveness of a brand – and thus to its competitiveness, success, and sustainability. The pandemic context confirms this: for a retailer, being able to set the right Price at the right moment is a matter of survival. Consequently, competitive pressure on Price is stronger than ever. A pricing strategy that is differentiating and attractive to consumers cannot today ignore competition. This is where effective product matching comes into play.
Competitive comparability as the pillar of pricing strategy
SETTING PRICES BASED ON YOUR COMPETITIVE SCOPE
The collection of competitive data – prices, descriptions, images, EAN – is the preliminary step to any product matching. For offline and food commerce, the data are generally acquired from panelists. They can also be collected online, by extracting data from competitors’ websites (Web Data Collect) or in stores (Scan In Store). If the raw data collected are, in their current state, of limited use, product matching allows for their valorization.
Indeed, matching consists of linking identical or comparable products sold by the brand and one or more of its competitors. To facilitate their interpretation, these links are organized in analysis tables. By going through them, the brand can compare each of its products – and their prices – to those of the competition.
The objective of this comparison is simple: to define the price positioning of its competitors and incorporate it into its future pricing strategies, using relevant and comprehensive competitor price data. The rules for creating pricing strategies can thus be customized based on the matched data.
Furthermore, the Matching module of the Optimix Pricing Analytics suite uses Artificial Intelligence.
ARTIFICIAL INTELLIGENCE AT THE SERVICE OF EFFECTIVE PRODUCT MATCHING
Although crucial in any competitive pricing strategy, the product matching process is generally tedious. It often requires the user to proceed with product by product chaining. Depending on the number of references and competitors of the brand, the task can quickly become time-consuming and discourage your teams.
To remedy this, the Matching module of Optimix calls on Artificial Intelligence at several levels. The module facilitates and speeds up the matching. It allows for two types of links: automatic chaining based on the EAN and chaining using AI to support the user.
Product matching automation based on the EAN
The automatic linking requires no human intervention. The algorithm analyzes the imported competitor data and compares it with the retailer’s product data. When it identifies identical EANs, it links the corresponding products together. This linking is automatic.
Obviously, this process allows our clients to save significant time and improve efficiency in product matching. It greatly contributes to generating a reliable and qualitative product database, useful for improving their competitiveness and pricing positioning.
Depending on the retailer’s industry, it is estimated that between 40 and 80% of its products can be matched in this way. However, automatic matching is not always possible. In these cases, it is up to the user to check that the products to be linked are indeed identical or comparable. In this case, AI supports their decision-making.
Manual linking in one click
Several scenarios require a user’s intervention, especially when the products are:
– Identical but their EANs do not match,
– Comparable or similar because they are from a private label,
– Identical but their size, content, volume, etc., do not match.
Again, Artificial Intelligence supports a simplified decision-making process and accelerates matching. Let’s see how.
The Matching module of Optimix Pricing Analytics allows for easy linking on two fronts: the photography of the products and their description. The goal: to match products in a single click.
For this, the AI compares the photographs and descriptions of products in the database. It normalizes the descriptions of competitor products using a synonym dictionary, then assigns them a proximity score weighted by prices. Based on this scoring, it sorts the competitor products and brings them closer to the most relevant retailer’s products. After that, the user only needs to determine if the products are identical, comparable, or comparable with a coefficient. Therefore, in one click, they can proceed with the matching.
Artificial Intelligence operates on the principle of machine learning. In concrete terms, this means that the more it contributes to the matching, the more efficient its analysis and scoring will become, further accelerating matching tasks.
Upon a user’s request, it is possible to set a threshold at which the AI automatically matches products. For example, we can define that the AI will match all products with a proximity score above 98%.
If, despite everything, the retailer hesitates to allocate high-value resources to the few remaining manual tasks, Optimix provides the necessary resources to outsource the matching.
Si, malgré tout, l’enseigne hésite à mobiliser des ressources à forte valeur ajoutée sur les rares tâches manuelles restantes, Optimix met à sa disposition les ressources nécessaires pour sous-traiter le matching.Â
A fully customizable reporting
The module has a dedicated space for the statistical analysis of the matching results. It includes, for example, data on the level of directness or the percentage of the brand’s products matched. The reporting is fully customizable to adapt to your scope and needs.
