Soft Solutions

ibs Analytics

THE ADVANCED DATA DECOMPOSITION MODEL (ADDM)

A sale is the result of the consumer perception of several Features such as: The banner strategy, which has a huge impact on the consumer behavior by defining prices, promotion policy and marketing. The local microclimate, induced by the concentration of competitors and the type of area (rural or downtown). The in-store availability and accessibility are also key features in a sale, with the assortment strategy and the planogram disposal.

• Some external factors are to be taken into consideration like unemployment, growth rate, inflation rate, which have an impact on the consumer purchase's budget. Moreover, household happiness can increase sales, as people are willing to buy.

“SALES ARE MAINLY RELATED TO CUSTOMER PERCEPTION OF VARIOUS COMPONENTS, SO IS THEN FORECAST”

Although some of theses factors are hardly measurable, some others can be detected and even decided by the retailer. Moreover, a sale can be defined by some item's specifics:

• General trend,
• Seasonal cycle,
• Seasonal peaks,
• Price changes,
• etc.

SALES FORECAST MODELING

From the sales history, ibs Analytics builds a forecasting engine of regular sales. By applying information mining & other methods, ibs Analytics extracts for each item several of its specifics components: General Trend - It models long-term variations of sales at the scale of the year Cyclic Trend - It models seasonal variations of sales at the scale of the quarter / month

• Weekly Variations- It models short-term variations of sales, represented by the variability of customers and changes in level of sales from a day to another
• Seasonal Peaks - It reveals repeating peaks in sales, mainly related to calendar and non-calendar events
• Baseline Forecast - It combines the above mentioned services, to provide a baseline forecast to end-user

MARKETING POLICY MANAGEMENT

In addition, to the above forecast results, providing expected sales under same hypothesis, the below models allow to reflect change of these hypothesis on the forecast sales level and provide required simulation for the various connected businesses (offer management, pricing, promotions, markdown, allocation…) Regular Elasticity - Increase in volumes depending on the level of price change

• Promotional Elasticity - Increase in volumes depending on the promotional policy (Single, BOGO, BAGB / Bundles)
• Elasticity Built Curve - Price change / promotion effect over time and its diffusion
• Flyer and Advertising Elasticity - Impact on communication over sales
• Halo and Cannibalization - Cross effects between items and sales correlations

Key Features