Methods of correct demand forecasting
How to plan assortment under stochastic demand?
While planning the assortment, chain retailers are facing the fact that, on the one hand, customer demand is random, and on the other hand, the retailer needs to control the availability of thousands of items on the shelves and in the warehouse. Because of this, optimizing the movement of goods from supplier to customer is very limited.
There are difficulties in forecasting customer demand. As demand forecasting is based on sales results, lack of goods on the shelf, low level of logistics and inventory management leads to distortion of information and false forecasting. Even more distortion of the overall vision is caused by over-sorting.
For the reasons above, retailers cannot:
There are difficulties in forecasting customer demand. As demand forecasting is based on sales results, lack of goods on the shelf, low level of logistics and inventory management leads to distortion of information and false forecasting. Even more distortion of the overall vision is caused by over-sorting.
For the reasons above, retailers cannot:
- Offer a range of products to match customer expectations;
- Optimize the supply chain of goods from supplier to warehouse and from warehouse to stores.
- The retailer does not know what assortment the customer actually needs and in what stores, because demand differs from store to store. Even if the right product is on the shelf, there is still a risk that the customer will not find it.
For assortment planning we usually use absolute revenue per unit of sales or profit from the category; or dynamic values - absolute and relative growth characteristics. It is important to keep the balance of categories, to make sure that some goods do not prevent the sale of others, and on the contrary, to create a chain of sales, when one product contributes to the sale of another.
It is very important to consider the selling space occupied by each product category, logistics costs and the business role of each category. Within each category, determine the optimal number of product items. To implement such a strategy, managers must see statistics on sales history, and at the beginning of the article we have already seen that the existing statistics often have significant errors. And with errors in statistics, all calculations and constructions based on statistics are meaningless.
All of the above problems can be easily solved with the help of the GreenShelf constructor. It also provides additional features:
It is very important to consider the selling space occupied by each product category, logistics costs and the business role of each category. Within each category, determine the optimal number of product items. To implement such a strategy, managers must see statistics on sales history, and at the beginning of the article we have already seen that the existing statistics often have significant errors. And with errors in statistics, all calculations and constructions based on statistics are meaningless.
All of the above problems can be easily solved with the help of the GreenShelf constructor. It also provides additional features:
So how do you properly plan your assortment?
- 1An accurate estimate of sales volume for each category
- 2An estimate of the relative share of sales of the product compared to the whole category
- 3An estimate of the frequency of purchases
In addition, the tool allows you to:
- 1Correctly assess the occupancy of the trade equipment for each item of goods
- 2To carry out comparative analysis of several products in dynamics
- 3Sales history and thus identifying seasonal trends
- 4to evaluate the result of sales in connection with the space occupied by it on the shelves
GreenShelf analyzes the demand of goods with a specific set of properties
The tool also allows you to use a modern approach to assortment planning, which consists in assigning a set of properties and attributes to each product. Demand analysis considers the demand not for specific goods, but for those or other properties that a number of goods possess. With this approach, it is possible to forecast the sales of new products that have a specific set of properties. In addition, it is possible to correctly substitute one product for another without losing demand.