By : James Nawrocki, Retail Software Solutions Consultant, Zebra Technologies

January retail sales are a seasonal headline grabber. Shoppers want the best bargains, while retailers want to know what their peers and competitors are up to and generally how sales are for the industry as a whole across food, fashion, electronics and cosmetics, among others.

Markdowns, discounts and special offers are a familiar approach to selling inventory that perhaps wasn’t in great demand among shoppers, or items were priced too high or wholesale order volumes were excessive. It’s not an ideal scenario, as retailers risk losing revenue and no one wants clothes going to waste and potentially ending up in landfill sites.

The traditional approach involves high-level demand forecasts, often calculated at the category or regional level. However, it’s no secret these models lack precision and fail to account for store-specific and hyper-local demand variations.

To maximise profitability and efficiency, an AI-powered pricing solution must go beyond static markdown schedules and basic demand projections. So, it’s important to look for solutions that offer granular demand forecasting at the SKU-store day level, advanced price elasticity modelling, and dynamic markdown optimisation.

It should also integrate seamlessly into ERP and planning systems since the AI is only as smart and helpful as the data it can access. AI needs unfettered access to clean, accurate data. Why are these things so important?

AI-powered pricing solutions forecast demand at the SKU-store-day level, allowing retailers to tailor pricing based on real-time sales trends, localised demand patterns, and inventory levels at each store, intelligently automated decisions thanks to better data asset visibility.

For example, a fashion retailer used AI to detect store-level demand trends for women's boots. Urban stores were selling out quickly, while suburban stores had excess inventory. Instead of blanket markdowns, the retailer optimised pricing by maintaining full-price sales in high-demand areas and discounting selectively in slower-moving stores. The result: the retailer delivered an 18% improvement in sell-through and a 9% increase in margin.

And when it comes pricing models, retailers rely on simple price elasticity formulas that fail to capture the full impact of price changes on different customer segments, product categories, and market conditions. But AI-powered pricing dynamically measures true price elasticity by incorporating historical sales data, historical promotion data, external data (competitor pricing, economic trends, and seasonality), stockout and anomaly detection.

One luxury handbag brand reduced unnecessary markdowns by implementing AI-driven price elasticity modelling. Instead of a flat 30% markdown at the end of each season, AI identified that a 15% discount on bestsellers and a 40% markdown on slower-moving SKUs would maximise revenue. The result: the brand had a 12% reduction in margin loss while accelerating inventory clearance.

For a $3B fashion department store retailer with nearly 30% of sales from clearance, changes in customer demand cycles along with a less than responsive supply chain left too much inventory in some places and too little in others.

Across the board, price cuts helped drive down total inventory levels but did nothing to address out of kilter assortments so that, in addition to the margin hit from ill-timed discounts, sales also dropped as customers were unable to find the items they wanted.

By implementing an AI forecast-advised markdown cadence, markdown recommendations were able to reflect true demand for seasonal items, by product and by location. In six months, margins increased four points and sell-through increased over 10 points.

Retailers will also leverage pre-set markdown calendars that reduce prices in fixed increments without regard for any demand signals or inventory levels e.g., 20% off after four weeks, 40% off after eight weeks.

But AI dynamically adjusts markdowns in real-time based on inventory levels, sales velocity, and external demand signals. A fashion retailer launched a holiday collection expecting strong demand. However, AI-driven analytics detected slower-than-expected sales in certain styles. Instead of deep discounts across the board, the AI recommended a 10% discount online and a 25% discount in underperforming stores. Thanks to its AI solution, the retailer drove a 15% margin improvement and a 30% sell-through increase.

If retailers do not have AI assisting them with pricing strategy, they are leaving money and margin on the table, as evidenced by the real life examples above and many other examples of companies who have leveraged AI’s fast modelling capabilities to increase their revenue and improve margins.

It’s also important to remember that not all AI pricing systems require a rip and replace of what’s already in place. There are modules that can augment an existing ERP and planning system. All the examples above were retailers who augmented with AI rather than replacing old systems with new ones.

Learn more about what AI can do for more dynamic sales pricing here.

Whatsapp