Ideally, you want to have at least two years of historical data for the most accurate and reliable modelling. This provides sufficient data to capture seasonal variations and trends in the model, ensuring a model that is more representative of real-life conditions. Additionally, for Brand tracking, we recommend having available data of several years, as this allows for a deeper understanding of brand performance in the long term.

The statistical model can identify uplift effects for marketing campaigns and other similar activities. In this way, we can give specific recommendations regarding budget allocation to achieve the most optimal results.

As a client, you will need to give Fieldstream access to the specific data to be used. This can be easily done via our many built-in data connectors, which seamlessly syncs data with the Fieldstream platform. In addition, we suggest providing us with company-specific knowledge and insights that may be important for analysis.

You can continuously compare model predictions with actual results, allowing for ongoing refinement. As more data is added over time, the model’s accuracy will improve.