Planning promotions using analysis of already accumulated data

[sadiksojib35]

When there is enough data and you want to conduct promotions regularly in different categories to meet the discount needs of different consumer groups, analytics will help you create a strategic promotional calendar for the year . In each promotional calendar period (usually a week), the retailer rotates the categories in the promotion, the products within the categories and the depth of the discount.

This approach is used, for example, by large grocery namibia whatsapp phone number retail chains - they have a catalog of discounted products for each week. But this is also the most difficult approach from an analytical point of view, since it is necessary to create a promotional calendar for a year in advance.

Since long-term planning at the product level is difficult, you can start by defining the category and duration of the campaign: for example, one week there are discounts on dairy products, the next - on snacks, then on household goods, etc.

As these weeks approach, you can move on to tactical planning - choosing specific products within categories and the depth of discounts. This helps to avoid the "promotional needle", when, for example, seeing a discount on "Domik v Derevne" one week, and on "Vesyoly Molochnik" the next, and so on, buyers will always wait for a promotion on milk and will no longer buy these products at the regular price and will not look at other brands.

Instead, you can first put a discount on "Domik v Derevne", then end the promotion on milk and start a campaign in the cosmetics category, etc. By alternating the promoted assortment, you can attract the attention of different groups of consumers, who will not get used to only promotional prices.



How else can AI be useful in planning promotions?
Finally, obtaining detailed analytics and evaluating the effectiveness of promotional campaigns also provides an opportunity to take the interaction of the retailer with the manufacturers of goods to a new level. Conducting a solid promotion with excellent results, effectively presented with the help of analytics, can become an important argument in business negotiations on purchase prices and retro bonuses - compensation that suppliers pay retailers for the successful promotion of their goods.



Expertise & AI
Many retailers conduct promotions based on expertise, when a specialist suggests a discount, considering it beneficial for the company. It is critical to supplement the subjective opinion of an expert with analytics and predictive models. Big data and AI models substantiate the parameters of the promotion, verify its potential and prevent experts from conducting promotional campaigns that will lead to a negative effect.

It’s hard to believe, but it’s precisely the stage of thorough analysis with the help of AI that still remains a weak point in promotion planning for most retailers around the world. Despite the fact that retail chains have a huge amount of data, very few implement AI algorithms specifically for promotion forecasting, relying only on the expertise of category management.

If you rely only on the sales growth indicator (which does not always lead to profit, as we have already seen) and do not take into account the criteria of “cannibalization”, bulk purchasing, purchase shifting, etc., you can lose a significant portion of your profit or get hooked on the “promotional needle”.

The transition to innovative promotion planning and business process change often comes up against experts who are not ready to rely on the help of AI and machine learning models. Retail staff should be willing to use big data and machine learning models, and not resist it for fear of losing their voice in promotion planning. Therefore, change management is an indispensable component of analytics implementation.

Promotions and discounts are important levers for managing sales and profits in retail, and their comprehensive planning helps to avoid the “promotional needle” and interact more effectively with suppliers. Analytics based on big data and AI are designed to help with this.