Promo is an effective tool for stimulating sales and a significant cost item for manufacturers and retailers. Promotion forecasting allows you to optimize the supply chain for promotional products and reduce business costs. How to get a high-quality forecast, and what pitfalls are encountered at different stages of the process, let’s figure it out.
A bit of promo theory
There are various promo mechanics: a discount on a product, 1 + 1, such as “buy one product, get the second as a gift” or when buying the first, the second — with a 50% discount, an offer to buy a set of goods or three identical units for the price of two or, buying for a certain amount, customers receive points, which can then be used for further purchases, and other options. The initiator of the action can be both suppliers offering product promotion, and chains, in order to stimulate demand and attract customers.
The ways of interacting with customers can be very different: a booklet, a catalog, mailing to the subscriber base through mail services, or instant messengers.
The chain will also have to determine the number of product positions that should be included in advertising on billboards, how many cells should be allocated for advertising in the catalog, how many shares should be from suppliers, the depth of the discount, and the relevance of the offer to the buyer. That is, the chain sets the rules for the promotion.
In the future, the retailer determines the timing of the promotion, agrees on prices, and then the question arises: how much goods should be purchased? There are usually three possible solutions:
- Purchase volume recommended by the supplier.
- The buyer determines the volume based on his previous experience.
- The recommended volume is calculated as part of forecasting and ordering (special systems are calculated based on analytical models).
What is very important to consider is the promo growth rate relative to regular sales. As one of the approaches to its calculations, the principle of price elasticity can be applied. The price elasticity of demand reflects the degree of buyer’s response to changes in the cost of the SKU. For example, a chain usually sells 10 bags of sugar. If the unit cost is reduced by 10%, we will get a 3 times increase in sales. With a discount of 20%, the increase will increase by 5 times. That is why, the sooner the chain agrees with the supplier on the promotion and receives a better price for the product, the more “gap” it will have for the maximum possible price elasticity. Therefore, it is more accurate to calculate growth rates from regular sales. Usually all calculations are based on the history of previous shares.
At this stage, possible means are used to calculate the volume for different types of promotions. For example, for a 1+1 promo, you need to calculate the number of additional SKU, predict and order it. Based on the results of the calculations, we obtain the predicted quantity of goods. In the future, the chain places an order for the warehouse and distributes the products to stores in accordance with the rules defined in the chain.
These rules depend on what levels of sales provide stores, their format and location. It is appropriate to note that there are promo-dependent stores in which goods are sold only if there is a promotional offer. Also, the distribution is highly dependent on geographical location. It is more likely that in a rural store, coffee sticks, even at a very favorable promo price, will remain unclaimed.
To summarize, when organizing a promotion, it is important to ensure the purchase of the required volume of products before the start of the promotion, carry out the distribution and then maintain its stock during the entire period of the offer. An important thing when predicting a promo is planning to exit the promotion. Let’s talk about all this in more detail.
How to start promo forecasting?
In any calculation, input data is important. What information needs to be collected and stored so that promo forecasting is as accurate as possible?
Firstly, you need to maintain a marketing calendar, or, in other words, a calendar of cause-and-effect relations, in which data on promotions are recorded. Let’s say a specific product participated in a promo with a 5% discount on certain weeks or days, and what real sales were as a result of the promotion.
Secondly, you need to classify all promotions, that is, break them down into types. Each type of shares is assigned its own identifier. As a result, the types of promotions are systematized.
Thirdly, the necessary statistics for predicting promo necessarily include historical data, which include:
- promotional sales history, providing insight into how SKUs were sold in the past. The company must maintain a promo calendar (fix promo periods, which SKUs and stores participated, what were the growth rates);
- the history of sales of goods constantly involved in the promotion. To translate / explain it, you can use a special variable — impact (impact). With the help of an impact, you can explain to the system an event/promo that has happened or happened in the past. That is, the impact shows the trend of SKU sales in past periods and suggests that the same result will be achieved in the future.
Also, when predicting promotions, it is necessary to take into account the impact of promotions on regular sales.
- Direct impact — when we take into account the history of promo sales in the past and can predict future demand.
- The indirect impact of a promo product on the sales of other SKUs can have absorption (cannibalization) or chain reaction (domino effect) effects. A simple example: sales of capsule powder at the regular price can be significantly reduced if the promotion includes weight powder at an attractive promotional price. This effect is also called cannibalization or demand switching. Impact can have the opposite domino effect (chain reaction) where a good promotional price spurs the purchase of another related SKU. The simplest and most successful «tandem» is beer and nuts/crackers.
As a result, based on all this data, it is possible to correctly calculate the forecast of the promotion for the future. Manually making calculations is labor-intensive and time-consuming, unlike a specialized system where the entire calculation is automated and provides high accuracy.
The mechanics of promo forecasting in the system is simplified as much as possible. The user only needs to select the stock for which an identifier (impact) was assigned at the previous stage. Next, you should define the product, store, week, for which the promo forecast will be calculated based on historical data and the impact that explains the sales trends in the system in the past.
If necessary, the forecasting system can take into account factors such as:
- additional display on the shelf;
- additional places of display;
- other factors affecting volume.
The resulting forecasting result will contain regular demand, seasonality, special periods and promotions. The entire calculation process is automated, the user always has the right to make adjustments.
Optimal distribution of the volume of promotional goods
Based on the results of forecasting, an order is formed, the total volume of goods is delivered to the warehouse. But how to properly distribute it among stores? For optimal distribution of inventory in an automated mode, C4R supplies a special solution based on the software of a well-known vendor. The system takes into account many factors, such as the distribution plan and keys: which store and in what priority, by what coefficient should a certain volume of goods be received under the promotion.
A key is essentially a distribution rule. It denotes, for example, store formats or other features of a retail outlet. If the store complies with this rule, then the consignment for distribution will be adjusted in accordance with it.
Without a dedicated chain forecasting solution, you have to rely on employee experience, store size, sales volume, and other store performance metrics. This approach often leads to errors, and those, in turn, to additional costs for the redistribution and transportation of goods between stores.
Exit promo the right way
The exit phase is the final stage. Its main task is to ensure that the retail chain exits the promotion with minimal inventory. This phase can begin at the very height of the promotional offer.
Important to consider:
- Phase start date: determined individually in each case. This is due to the specifics of the SKU, logistics for its redistribution, etc. The time to exit the phase depends on the history and seasonality of sales. For example, to speed up and stimulate sales of fresh products, a deeper discount should be given by the end of the promo.
- Sales dynamics: this helps to redistribute the purchased volume of products in time. If a product is not sold in one store or group of outlets, then it can be moved to outlets where the sale is faster.
Promo forecasting is inextricably linked to the distribution of inventory throughout the supply chain. In turn, the correct forecast allows you to reduce the cost of logistics and maintenance of residues after the campaign. On the scale of the trading chain, these are large budgets that affect the margins of the business. Chain management is looking for ways to optimize such costs.