Context and objective
The client is a consumer goods company which products are sold for on-premise consumption by hundreds of thousands of retailers: a very granular and diverse retailer base
Objective is to gather ‘deep data’ about retailers in order to assign them to specific segments and allocate sales and marketing resources to the accounts with business potential
Framing of the segmentation and data sources identification
Segments have been identified by marketers based on their knowledge of the market (incl. market analysis and surveys). All retailers should be assigned to a segment based on:
Some individual, structured and quantitative data are available from the company’s CRM. However, some criteria require data that are less structured and not available internally to be gathered and interpreted. In particular, data for the image criteria seem to be available from digital platforms (ratings, price information, dress code information…) while data for consumer typology seem to be available from social media).
Data scrapping and big data analysis
Emerton has built tools to automatically and dynamically scrap data from the external sources, via API calls and robots. Those data are then fed to big data algorithms in order to transform them into structured and usable data.
The resulting big data table can in turn be used for segmentation.
In order to refine and validate the algorithms (for interpretation and segmentation), a sample of retailers is used as a test sample: real-life observation enables to properly allocate them into the segments; the comparison with the big data analysis allows to identify discrepancies and adjust the algorithm accordingly.
Outcome: ready-to-use retailers segmentation, and re-usable algorithms
As a deliverable of the project, the allocation into segments of thousands of retailers allows for an in-depth analysis by segment:
In addition, the tools being dynamic and largely automatic allows for an easy re-use or update.