Situation
The vision of a global Home Furnishing player is to create a better everyday life for many people. It is the key for our client to understand the competitive landscape and investigate how to distinguish itself from its competitors. Our client is looking for ways to match its assortment to competitor assortments in an effective and efficient way.
Structurally researching all the markets they operate in (>50), across all the product areas they offer (>200) and comparing against multiple competitors per market is a monumental task. The current approach has a number of key issues; it is too labour-intensive and thus expensive to scale, the quality of results are low and the solution is an external black box that doesn’t learn over time.
Approach
Together with our clients experts, we created an in-house solution that uses AI and machine learning to handle heavy workloads. We used machine vision to compare and match product images, natural language processing to compare and match product descriptions, dimensions and materials as well as machine learning algorithms to identify similarity.
Results
The solution is scalable to all markets, learns and improves over time due to validations by product, and already provides high-quality results using (just) the power of AI. This enables our client to use insights to adjust pricing strategies, identify product categories where there is a need for more efficient design and closely monitor market movements in the dynamic retail environment.
