Forecast of production needs

Application Track:

Code:

Domain:

Proposed by:

Entity Logo:

Summary of the entity:

Grupo AN is a centenary cooperative and a leader in the Spanish agri-food sector. We are a second-degree cooperative formed by 160 agricultural cooperatives and 40,000 farmers and livestock owners.

Grupo AN is the largest producer of cereal in Spain and is also highly active on national and international wheat, barley, corn, rape and sunflower markets. We are mainly producers of agricultural products, including fruits and vegetables of worldwide commercial interest. We also offer the consumer our products in all possible formats: fresh, processed, fresh cut, refrigerated and canned.

Grupo AN is also one of the largest operators on the poultry market and an indisputable leader in the supplier market, Grupo AN commercializes fertilizers and fuel, which is distributed in Spain through a network of petrol stations, which are opened to the public.

Apart from that Grupo AN also has a network of Processing companies.

One of them is the Industry of Vegetable Preserves that prepares canned products which are ready to be consumed all year round. Products are commercialized both in national and international markets.

In addition, Grupo AN has a fourth and fifth range processing company. Under the commercial brand ‘Diquesí’ Grupo AN offers its products, prepared at its factory in Tudela.

Summary of the challenge:

The challenge is creating a platform for predicting customer orders. Currently our companies receive orders from their customers with very little notice (usually on the same day, except for large orders that arrive 2 days in advance).

At this time, our companies have a specialized staff for production planning and they make decisions according to the historical orders and their market information (through contacts with clients and updates from social networks). 

The problem is that when the specialist is out of the office, it is difficult to make forecasts correctly, since it is knowledge that cannot be recorded.

Data:

1. AN Database with historical orders (quantity, product, price, sales, company):

• COMMERCIAL NAME (TEXT): name of the buyer
• FISCAL NAME (TEXT): name of the company
• PLATFORM (TEXT): place of shipping
• FAMILY (TEXT): product name
• ARTICLE CODE (TEXT): product code
• ARTICLE (TEST): description of the product
• WEIGHT (NUMBER): weight of the package
• UNITS/BOX (NUMBER): number of products in the package
• BOX/PALLET (NUMBER): number of boxes per pallet
• SALES BOXES (NUMBER): number of boxes sold
• SALES UNITS (NUMBER): number of units sold
• SALES KG (NUMBER): number of kg sold
• SALES € (NUMBER): price of product sold
• LOGISTICS €/PALLET (NUMBER): cost of each pallet
• LOGISTICS € (NUMBER): total cost of logistics

2. Market info (clients, social networks …)

3. Generic market studies of the sector (Nielsen, Kantar)

4. Other variables that could be integrated:
– Compare competitive offers
See how these variables influence customer demands: price, weather, beginning-end of the month, sale city-rural area …

Expected outcomes:

To create a platform for predicting customer orders. We are looking for a non-static tool.  We need a tool that can be constantly fed with new information.