Posted on 2021-08-03
Ryax Technologies Ryax Technologies

Get a personalized sales forecast for all your items

+ More details

Description

Not satisfied with your sales predictions on Excel? Indeed. Predicting sales isn't linear nor a simple moving average. This forecast analysis will provide you with an overview of the predicted sales for every item of your catalog - it does take seasonality into account to measure the potential to be sold in a defined timeline.

We recommend to perform this analyis in a timeline matching the product lifecycle, for example, at the stock arrival date. It can either be an automated analysis sent periodically through mail or a regular process pushed to your favorite supply chain ERP/demand planning tool.

Analytics status

  • Available

Business benefit

The major positive impacts to be expected by running this analysis are essentially on :
▪ total turnover
▪ sell-through rate

Data inputs (mandatory)

We'll be using historical sales data containing the following information :
▪ Date : historical timestep of observation (could be daily, monthly, weekly…)
▪ SKU :SKU ID, item identification number
▪ Revenue : sales in quantity or revenue

Data Output

You get two deliverables :
▪ the predicted sales per SKU
▪ a graphical presentation of the predictions

Technical description

We use a regular ARIMA model to deliver this prediction. ARIMA refers to auto-regressive "AR", integrated "I", and moving average"MA". This is a statistical analysis model that uses time series data to either better understand the dataset or to predict future trends.

Limitations

This model does not take into account product cannibalization or sales pattern similarity (sold simultaneously, correlated products)