Forecasts are typically produced at the lowest level (per product or per SKU) because that's the data aggregation level your company needs to optimize its inventory.
Multi-sites, multi-channels or massive numbers of SKU are natively supported, as our technology is cloud-native and has been designed to handle an arbitrarily large amount of data (Lokad has received the first Windows Azure Award in 2010). In particular, Salescast can deliver forecasts at the Point of Sales level (even for large retail networks) or at the warehouse level.
Results can also be exported by Salescast directly to your existing IT systems, typically into an SQL database, feeding further automated processing. Excel reports are handy for direct consumption, but re-importing our forecasts from an Excel sheet would be impractical.
Refreshing your report takes 1 click, and can be done anytime. The refresh can even be scheduled and automated. Once refreshing is triggered, it typically takes less than 60 min for Salescast to complete the operation. Since Lokad relies on a powerful computing-intensive forecasting algorithm, we can't deliver results in real time.
Cyclicities, as a generalized notion of seasonality, with recurrent demand patterns within a year (classical seasonality), within a month (paycheck effect) or within a week (day of week effect).
Demand disruptions, such as promotions, stock-outs, outliers, such as uniquely large bulk orders and other business events.
Calendar artifacts such as public holidays and quasi-seasonal effects (Mother's day, Easter, Ramadan, etc.).
Intermittent products (e.g. low sales volume).
And many more.
Most of the forecasting features are not specific of Salescast, but only reflect our core underlying forecasting technology (see also the FAQ for details about patterns natively supported).
Yet, with our fully managed forecasting service, nobody in your company needs any skills in statistics to benefit from all these features. There are 2 and only 2 forecast settings: the period (day, week or month) and the horizon (how many days/weeks/month ahead you want to be forecasted).