I can report from the experience of many projects: Markdown Optimization is one of the topics that retailers are particularly concerned about – across all branches. Because markdowns – and especially write-downs – limit gross profit. Therefore, it is important that retailers proactively manage their sales and keep write-downs as low as possible. This is not least essential for the overall success of a company. The particular challenge with markdown pricing is, that those products whose prices need to be intelligently reduced, are already at the end of their product life cycle. As a rule, the following applies to such products:
- the demand for these items is constantly decreasing and customers expect reductions
- Pricing managers lack the time to price these items in a targeted manner and keep losses low
- these items cost storage capacity and sometimes cause additional disposal costs
For pricing managers, these special circumstances create a real bottleneck: in order not to sacrifice margin needlessly, they would have to regularly determine which items of their entire assortment meet these criteria, and for those that do, what the appropriate price reduction for each one should be. This would take far too much time. In the end, a “spray and pray” approach to reductions are generally accepted as a necessary evil, even though they are strongly detrimental to margins. It’s akin to a watering can – only some water willreach plants, and the rest will be wasted on the ground. You can hardly earn margin with such a watering can approach; a more direct nozzle is needed.
From an economic point of view, it is essential to rethink the pricing process for the markdown optimization. The following objectives must be combined:
- Daily price adjustments based on demand to achieve margin-saving sales (= abolition of the watering can)
- Adherence to all sales targets and optimal use of storage capacity
- Streamlining of processes and reduction of manual effort
Fueled by AI, these solutions can forecast sales per product for the coming days based on the latest transaction data (current clicks, created shopping baskets, purchases, etc.). Depending on the specified target sales date, the daily price is then determined for each item. If the target sales date is close and the sales forecast is poor, the price is reduced. If the sales forecast is good, the price is either minimally reduced or is even increased.
In contrast to the watering can approach, an algorithm re-evaluates each individual product every single day – making the top priority the retailer-specifiedsales date. Measured against this, the algorithm only reduces the price of the products that would not be sold by the target date. All other products are either only very gently discounted or – if there is demand – the price is even increased again.
Across all products, you can then create a profitable balance between aggressive selling and consistent utilization of your margin potential. The sales process is thus margin-friendly, your goods go out of stock on a predictable timetable and you reduce your markdowns considerably.
The best thing is: the manual effort is remarkably low. How low? You should see it live!
