In the rapidly evolving landscape of the fashion industry, maintaining a competitive edge requires innovative approaches beyond traditional practices. AI-powered markdown optimisation emerges as a critical solution, enabling retailers to navigate dynamic pricing challenges and enhance profitability in an increasingly complex economic landscape.

The integration of Artificial Intelligence (AI) and Big Data has become indispensable for brands aiming to stay ahead by aligning with emergent trends and facilitating data-driven decision-making. Dynamic pricing represents a significant challenge within the fashion retail sector, particularly during critical sales periods.

The decision-making process involves a multifaceted analysis encompassing gross margins, inventory levels, and market dynamics. The increasing complexity of stock flows and the growing prevalence of high return rates add layers of complexity to inventory management, impacting the ability to set optimal discounts. 

The year 2024 has introduced additional challenges, marked by rising inflation rates, escalating trade tensions and supply chain disruptions, each contributing to unpredictable shifts in consumer purchasing behaviour. These developments render traditional discounting strategies less effective, highlighting the need for more sophisticated, adaptable approaches. 

In response, leading fashion retailers are turning to AI to transcend traditional limitations. By harnessing the power of machine learning and predictive analytics, these forerunners can distill actionable insights from vast datasets, enabling more nuanced and effective markdown strategies.

Using AI in Markdown Optimisation 

The integration of advanced AI models into retail operations marks a significant evolution in the fashion industry. This shift is particularly impactful in the realm of markdown optimisation, a critical area for fashion retailers seeking to enhance seasonal sales performance and maximising profits while minimising waste and reducing redundancies.  

By leveraging AI to optimise discount levels, retailers can achieve higher sell-through rates, effectively reducing excess stock and contributing to sustainability goals. This alignment with eco-conscious values is particularly pertinent in today's market, where consumer purchasing decisions are increasingly influenced by ethical and environmental considerations. 

In the context of the current economic landscape, AI's capability to analyse vast datasets and predict consumer response to pricing adjustments is invaluable. Retailers now recognise the critical importance of deploying the right AI tools to refine their markdown strategies, ensuring that each product is priced optimally to balance demand generation with profitability. 

However, the adoption of AI in markdown optimisation is not without its challenges. Retailers must navigate the complexities of selecting and implementing the most effective AI solutions, weighing the potential benefits against the costs and integration hurdles. Despite these challenges, the strategic value of AI in this domain is clear, prompting a careful evaluation of different tools and approaches. 

Having identified the relevance of this strategic business opportunity that is presented to fashion retailers, Metyis has developed an innovative methodology for building custom Markdown Optimisation systems, capable of handling the ever-changing landscape of the fashion retail business while also being flexible enough to be integrated into any client’s existing technological infrastructure. It is a tried and tested methodology that has already been implemented for European premium fashion brands. 

Metyis' innovative approach to markdown optimisation 

While Markdown Optimisation is not a novel business concept, traditional approaches to it are based on methodologies and practices that, despite being effective in a bygone era, are insufficient to compete in the current business landscape.  

A best-in-class Markdown Optimisation system requires not only a formidable forecasting capacity but also the flexibility to adapt to the specific commercial realities of each client. Additionally, it must be capable of generating a comprehensive range of scenarios that clients might encounter when considering different discount options. These essential aspects render our approach avant-garde. An exploration into the features of our tool uncovers even more interesting elements. 

There are five key areas that summarise the relevance and utility of Metyis’ Markdown Optimisation. 

1. Calculating price elasticities with the help of AI 

Calculating price elasticities is key to unleashing the strategic capabilities of the Markdown Optimisation tool. Measuring how clients react to changes in prices helps us simulate different scenarios and arrive at optimal price points. We leverage an AI similarity tool, which uses computer vision algorithms to match similar looking products together.

This tool uses transformer architecture to identify the most relevant qualities of new products in the assortment and maps them to the most similar ones we have sold in the past. By doing this, we can assign elasticities of similar products of the past to current products in the assortment, helping us make the forecasts and simulations much more accurate 

2. Predicting demand in all possible scenarios 

Our solution provides a sophisticated demand prediction model that caters to a wide array of scenarios. By integrating seasonal trends captured through AI models, baseline demand of every item and the product-specific price elasticities, we can get a dynamic and nuanced picture of potential demand over a wide array of price points. This detailed insight allows retailers to make strategic decisions about pricing and markdowns, maximising profitability while maintaining customer satisfaction.

3. Capturing stock flows 

By harnessing real-time data and advanced analytics, our solution provides an in-depth understanding of inventory flows. This feature can help retailers pinpoint patterns and trends in their inventory levels, facilitating more efficient and cost-effective inventory management. From restocking best sellers to marking down slow-moving items, this insight enables businesses to maintain a balanced and profitable inventory.

4. Retrieving optimal discounts for all products 

Finally, our solution is designed to calculate the optimal markdown for each product. By considering factors like initial cost, sales velocity, current inventory levels, returns and predicted demand, it can recommend the most profitable discount for each item, in each sales cut. This feature not only increases gross margins but also minimises leftover stock, contributing to a more sustainable and efficient retail operation.

5. Measuring the Impact  

After the conclusion of the sales season, it is of utmost importance to measure the impact generated by the recommended discounts applied, so we can gather as much information and feedback as possible to develop the next iteration of the solution. Metyis’ has successfully worked with clients in the Luxury and Premium apparel sector, and the Markdown Optimisation model has managed to achieve up to 4% uplift in gross margins, while maintaining or even increasing the sellout rate of their products by up to 10%.

To put this in perspective, this means that only by way of surgically optimising the discounts can we drive up sales without sacrificing margins, immediately making the business more profitable and competitive.

By creating this end-to-end solution, retailers can make the most of the data already gathered in their day-to-day processes and transform it into a powerful tool for devising markdown strategies. The end-user dashboard that Metyis delivers as part of this package, which allows exploring the discount recommendations and navigating client-specific KPIs, makes Markdown Optimisation a complete and compelling product for any retailer seeking to optimise their operations.

AI-based markdown optimisation is the future of promotion management for fashion 

Markdown optimisation is no longer an option; it is a necessity for fashion apparel retailers seeking to thrive in a dynamic and challenging industry. The integration of AI-powered models, considering stock, sellout, margin, returns, price elasticity, demand forecasting, multiple discount combinations, and product matching, empowers retailers with actionable insights that lead to increased profitability, reduced waste, and improved customer experience.

As AI technology continues to evolve, markdown optimisation will play an even more critical role in shaping the future of fashion retail, paving the way for a more agile, sustainable, and customer-centric industry. 

About the authors behind the article 
Arshiya Nagi is a Director in the Amsterdam office; Arpit Gupta is a Principal in the Madrid office; and Guillermo Tovar is an Associate in the Madrid office.