Markdown Optimization Helps Apparel Retailers Maintain Gross Profit and Profit Margin
Apparel retail is infamous for extremely short collection cycles. This means that businesses are always pressured to get rid of old or surplus inventory to make room for new collections. In this time-sensitive scenario, retail teams usually prefer to clear off shelves at all costs.
In many cases, retailers launch promos which attract customers and boost sales, but lead to profit margin losses. In fact, as much as 72 percent of promotional campaigns do not break even. This happens as pricing managers lack time to calculate optimal markdown prices for every discounted item, and instead set standard markdown prices.
However, such an approach fails to consider the elasticity of price — i.e., it doesn't allow managers to analyze how this particular item will react to price changes. For example, whether it's safe to sell a product at a slightly higher price than a standard markdown price or if it's better to offer an even deeper discount to raise the sales of other items.
Processing billions of data points, revealing the cross-impact and crafting the optimal offer in real time is next to impossible for humans. That’s why advanced companies are incorporating sophisticated pricing software based on machine learning to help managers hit three goals: clear off stock in time, earn as much as possible, and speed up repricing.
Machines can analyze infinite amounts of data and detect all the patterns humans can’t see, as well as suggest the best way to reach a set goal any time. Some retailers would argue that algorithms would offer to take illogical steps in order to hit a particular target — e.g., to sell a product at a 70 percent discount, which is a much lower price than agreed with a manufacturer. Meanwhile, price optimization self-learning algorithms are always limited by the retailer’s constraints and can be corrected any time.
Intertop, an apparel omnichannel retailer with 114 brick-and-mortar stores, partnered with retail pricing platform Competera to sell its old stock off and give way to new collections. Another goal was to do it while maintaining the gross profit and profit margin. Intertop switched from blanket (standard) discounts to differentiated promotions. Retail managers applied machine-generated, elasticity-based markdown suggestions for a group of discounted products for their weekly repricing cycles. As a result, the company saw a 200 basis points uplift in profit margin and a 10.3 percent boost in gross profit. All the selected items were sold on time.
Ilona Baskova, brand manager responsible for pricing at Intertop, commented on the efficiency of the solution: “Technology does boost the financial performance of your company. When using machine learning in repricing, we do not do repricing per se. We set the rules of the game and control the results. Machines do the rest of the job.” Also, using machines for routine pricing tasks lets pricing managers save up to four hours per person per repricing cycle.
The pricing tool integration required Intertop’s team to prepare historical data, course-correct algorithms, and be psychologically ready for a next-gen solution. “We were skeptical at first. I wouldn’t call it distrust, but still … I calculated new prices for the first repricing cycle myself," says Baskova.
“Then I compared them with Competera’s price suggestions. For the most part, we had identical prices. It seemed weird to me: how could it know what I wanted to do? After that, we just trusted the machine (and course-corrected, if necessary). And it worked!”
Intertop is ready to scale the solution to optimize pricing for its upcoming collection.
Related story: 6 Tips for Retailers Battling the ‘Amazon Effect’
Vladimir Kuchkanov is a pricing solutions architect at Competera, pricing software for online and omnichannel retailers. He is a Data Scientist, a top-rated domain expert in business analytics, pricing and media management with a successful track record in world-class FMCG companies.