Running a very large retail network on a smartphone
Long before ‘big data’ became the technology buzzword of 2012, retail networks have been among the pioneers in dealing with the large amounts of data that is produced by their supply chain and their point of sale systems. Recognizing the richness and immense value of their data, heavy IT infrastructure investments have, in many cases, been made.
However, to date, the limitations and cost of the required infrastructure has left the reality far behind ambition and promise. This is particularly true for the richest retail data source, which also dwarfs all others in size: receipts generated by point of sale systems. Collecting and processing receipts of hundreds or even thousands of stores has remained a daunting, and very expensive, task.
What about running a large retail network on a smartphone instead?
While this question is provocative both from a technical and commercial point of view, we explain in this whitepaper how fundamental operations such as collecting and processing receipts for retail networks of up to 1000 stores can be done on a smartphone. The source code used by Lokad to produce the results exposed in this white paper has been made available as open source under a very liberal license (BSD) on GitHub.
By sharing some insights on big data for retail, we hope to further fuel the advance in retail data exploitation. Any feedback will be appreciated, don’t hesitate to contact us.
