DO wanted to modify its DDE solution to require less historical sales
data, as specific clients didn’t have this. Due to limited datasets, the
existing statistical model could not produce accurate forecasting
outputs/expected sales. Field sales agents require the data to direct
them to the highest value opportunities at a store level to generate
incremental sales/return on investment for clients.
To make it a reality, DO opted to partner with Argility and leverage the
power of the Google Cloud Platform. Argility’s data team of experts
helped design a solution to achieve more accuracy in forecasting data
by shifting from DO’s traditional statistical models to a machine
learning (AI) algorithm. GCP Solutions utilised: Vertex AI, BQ,
Dataproc, Cloud ML, GCS and Compute Engine.
Including forecasting in the DDE solution resulted in improved and
more precise sales forecasts produced on October 31, 2023. This
enabled more accurate analysis of the incremental return on
investment associated with in-store initiatives, incl. more precise alerts
to trigger interventions where lost sales occurred. In return, this will
broaden the customer base and give a competitive advantage.
