3 assortment scenarios: examples of company sales forecasts

Artificial intelligence or AI is now at the service of the company. Faced with new forms of supply and the management of stocks in sometimes tight flow, the company must absolutely anticipate its sales in order not to fall out of stock and preserve its margins. It must also have a wide enough assortment to reach every customer likely to buy one of its products. There is now an assortment optimization method that uses artificial intelligence to create assortment scenarios for sales forecasting. Let's find out what an assortment scenario is in the commercial management of the company.

Definition of the assortment

A company's product assortment is composed of items sold in a physical store or online. To create its assortment is therefore to gather its products by category or by range and to offer them for sale, both to customers and to prospects or leads. The orientation of the assortment is thus one of the components of the commercial policy and strategy of the company:

• position on a given sector: public, professional;
• level of range : luxury, discount;
• catchment area: creation of the assortment according to the geographical area;
• specialization of the assortment: sport, leisure, technical range;
• distribution level: manufacturer, wholesaler, distributor, reseller.

his assortment is therefore the object, in addition to the financial investment it implies, of a precise study of the consumer by the implementation of a sales forecast scenario. This scenario will guide the company in the choice of its assortment. It will be assisted in the implementation of the commercial evolution towards which it wishes to be directed (continuation of the same target or reorientation of this one).

The choice of the assortment is logically based on the future customers. The company will thus carry out a market study which will be based on previous sales, loyal customers and the evaluation of prospects to convert. The market study puts forward different scenarios to follow, which will influence the choice of the company's sales range. The assortment is therefore a strategic point, both for the physical or online point of sale of the company, and in terms of commercial results.

Examples of artificial intelligence applied to the assortment

Before giving you examples of assortment scenarios, let's take a look at the arrival of artificial intelligence in the predictive management of the company's commercial actions. A classic company uses its old data (year n-1 and previous years) to create its future sales. These studies are generally carried out on the following basis:

• the amount of sales broken down by service or product line;
• the number of visitors online or in store;
• the average basket;
• sectorization by age group;
• consideration of particular times of the year.

This study, even if it is computerized, requires time and many business management skills, such as the handling of often complex dashboards. Whether you are interested in prospecting for new customers or in an existing customer base, these resources are still limited.

Artificial intelligence is revolutionizing the way companies carry out their assortment. It has the ability to learn from data embedded in the computer. This is called machine learning. To simply present the artificial intelligence, the company enters in the machine all its data which will be analyzed and learned by the algorithm. The algorithm will then understand the company's objectives and put them in relation with the learned data. It will then make cross sales forecasts from the product assortment and the company's development needs.

This is a far cry from traditional market research, since the power of artificial intelligence is deployed to serve the company's various acquisition channels.

Example 1: Forecasting promotional sales

The assortment scenario is concerned with forecasting promotional sales. A company may decide to buy a particular stock that will eventually be used for promotional sales in order to strengthen its cash flow at a given time of the year. Promotional sales can also apply to an assortment of products that will be optimized prior to the launch of the campaign.

The predictive model proposes, both for the articles and for the different sales channels (email prospecting, sales tunnel, etc.), different promotional scenarios that allow to anticipate and define the best marketing and sales actions to implement. The detailed analysis of all the data thus includes, without any possible error, all the elements that will make the promotional campaign a success.

The company's commercial strategy is based, in part, on the upcoming promotions. If the first goal of the promotional sale is to reduce the stock of certain products at the end of their life, it must not become a money-losing machine for the company. The promotion must remain a way to optimize the overall profitability of the company, despite a decrease in sales margins. The AI platform delivers a set of predictions composed of:

• marketing actions to be carried out: sending emails, conversion tunnels, print or online advertising;estimation of the best purchase price for items to be sold on promotion;

• selection of the assortment (range or product) to be promoted: articles to be destocked, new product to be promoted;

• Customer and prospect interest in the products: list of popular items, number of visits, shopping cart abandonment.

AI automatically calculates the impacts of the promotional campaign: quantities sold, turnover over the given period, commercial margins, stock level after the campaign, etc. It thus accompanies the decision before the launch of a promotion.

Example 2: Assortment scenario and product selection

The products composing the assortment are the object of a particular study, which must lead the company to have:

• a product adapted to the customer;
• a sales price in line with the market and the desired margin;
• a physical location of the product optimized to increase sales.

• a problem of margin calculation;
• a difficulty in increasing market share;
• the identification of certain products not to be renewed;
• an optimized choice of product variations;
• the storage capacity;
• an internal competition between two products.

These examples show us that creating an assortment scenario involves a sometimes complicated selection of products. Dealing with different brands or integrating new products into the range, without knowing how successful they might be with customers, does not make order management any easier.

Example 3: Predictions of consumer goods purchases

A company that specializes in the sale of consumer goods can use assortment scenarios in its procurement. Artificial intelligence can be used to create a specific product unit or range for customers and prospects.

Consumer goods range from furniture to sporting goods. Some structures may specialize in a given business sector or, on the contrary, sell products that are not related to each other. Procurement can therefore be totally different from one department to another and enter into a unique purchasing forecasting study for each sector of the company.

Artificial intelligence thus comes to the aid of each department. It has the ability to integrate all company information and perform a sequence of complex mathematical calculations on a specific sector of the company, through partial data extraction. This means that the use of AI within the supply chain segments the needs of each unit and provides predictions on future assortments through predictive examples of:

• moving upmarket;
• the acquisition of new sales channels;
• the elimination of a part of the range;
• the increase in investment resources;
• the calculation of sales price and associated margin.

It also allows to compare products between them and to check the absence of internal competition (cannibalism) which could harm the competitiveness of the company and be detrimental to future sales.

The assortment scenarios also make it possible to highlight possible stock malfunctions. This can be observed in the case of products being tied up for too long or overstocked, which directly impacts the company's cash flow, especially in the case of high value-added products. Through forecasting methods integrating the notion of sales and stock releases in real time, AI brings a certain serenity as to the triggering of future supply and also anticipates any stock shortage.

To conclude, the predictive information related to the assortment policy of the company is today an asset in its development. This tool allows a total anticipation as for the choice of the various articles, such as:

• precise determination of the range;
• the selection of the best-selling sizes;
• the quantities to be put in stock;
• the real needs of customers and prospects.

Artificial intelligence puts all its computing power and its incredible predictive capacity at the service of the company. Today, it is becoming a true management partner and can be adapted to each sector of the economy.