One of the top three market research companies operating in 90+ countries worldwide needed a better service pipeline for their retail display auditing projects due to increasing pressure from their client in Turkey, the market leader in the soft drinks category.
The existing process of the market research company consisted of manual counting of similar SKU groups in the shelves during a store visit, followed by sampling-based consistency checks in the back office. In the best case, the SKU count report of the store was made available to their client in two days after the visit. Results were not as accurate and detailed as required due to manual counting of SKU groups (rather than counting unique SKUs), not as consistent as necessary due to sampling-based back-office inspection (rather than 100% inspection), and obviously not as fast as demanded by the client. The aspiration of the end client in the soft drinks, and hence the market research company, for “Better” meant into three criteria:
The market research company came to us and asked whether Vispera Image Recognition Service can meet these needs. In order to demonstrate it was indeed the case, we set up a pilot comprising 270 modern channel store visits and 1400+ SKU covering not only the SKUs of their client but also of their competitors. In order to facilitate in-store photo collection, we provided our easy-to-use mobile app with modern look-and-feel, which can also be used to collect other auxiliary (non-photo) data in the store. The results of the pilot delivered our promise with over 99% accuracy with SKU-level detail and shelf position information for 1400+ soft drinks SKUs. The average store visit was reduced from 30 minutes down to 17-18 minutes, that is, by 40%. Reports were delivered within the hour right after the completion of the store visit via our web platform.
Groups of lead SKUs
Inaccurate count data
Reports in two days (best case)
1400+ unique SKUs
Reports within the hour