SkuCaster’s forecast accuracy is calculated a on a 3 month-forward forecast against actual data.

SkuCaster’s data science team trains an AI model on a minimum of 2-3 years of historical Sales, Shipment, and industry relevant data, excluding COVID-19 years.

During the initial model development, our team interviews your key stakeholders to clean and validate your company data so that the forecast incorporates specific information about specific SKUs.

Yes! Our SkuCaster algorithm has Characteristic Based Similarity “Like-Product” function that analyzes patterns amongst the most similar SKUs and Products to make the most accurate prediction of the future.

Ultimately, the SkuCaster forecasting engine depends on the quality of data. Our algorithms can choose between the most appropriate data and untrustworthy data and then matching to appropriate product types. Ultimately, if our team deems your data too untrustworthy, we will recommend you don’t use SkuCaster.

Weekly, Bi-Weekly, Monthly, Semi-Annually, or Annually. The more frequent the better because SkuCaster has more data to learn and iterate on.

SkuCaster provides performance warranties in its maintenance & support contracts. You can purchase an enterprise license agreement (ELA) or annual subscription with a one-time development fee.