Recent reports have highlighted the challenges many retailers are having with excess inventory. With supply chain disruptions continuing, purchasing more stock in advance to meet the forecasted upturn in demand made perfect sense. However, the boost in sales was short-lived as inflation and political instability took hold. This has resulted in a number of profit warnings as retailers have been left with too much unsold merchandise, resulting in knock-on impacts on cash flow, warehouse space and an inability to bring in new items. The impact of excess inventory at this challenging time can be devastating as seen by the recent demise of made.com. So, what can retailers do about this?

For many retailers, the first thought to reduce unsold items is to simply buy better. However, this is not a complete strategy. In these ever-increasingly fluctuating times, predicting the trends and planning the right assortment is a challenge. Forecasting ideal buy quantities is near-impossible. For many retailers, their methodologies are too simplistic. Using previous history never was valid but it certainly is even more unreliable today. Out-dated approaches based on sell-through do not help understand risk. And whilst more advanced approaches utilizing machine-learning can optimize this process, as Mike Tyson said, “Everyone has a plan until they get punched in the mouth.” Real-life kicks in, trends change and alongside death and taxes, excess inventory is certain.

For many retailers, this can equate to 20% of their purchases. In a recent poll undertaken by Blue Yonder, we found that the average merchandiser spent around half a day planning their markdown. Given its impact on profits, this is woefully insufficient. Understanding how best to manage end-of-life products is complex; there are many questions like how many units do you have, where it is located, how has it performed, what is the price elasticity, what cannibalization will there be if the price is changed, and what are the consumer value-drivers?

Uber and Amazon have changed the way customers shop, pay and engage with demand-based pricing. These apps alongside e-commerce websites have increased the visibility of pricing and consumers expect transparency. Consumers are certainly looking for value at the moment, but value does not just mean the price.

In the recent McKinsey report “Navigating inflation in retail: Six actions for retailers,” one of the key actions was to go granular with pricing and promotion and tailor value delivery to consumers. Going granular means understanding the different dynamics for each item at each store. There may be 5,000 units left of an item but the right price to clear will be completely different depending on whether all the merchandise is in your warehouse, a small selection of stores, or spread evenly across many. It will differ by product type, what similar items are being priced at, and external factors such as the weather and events.

In addition, what constitutes the right price? Is it the one that maximizes sales, margin, or that will clear the inventory quickest? When is the best date to make the price reduction or maybe it is best to have a small promotion on the right items running up to the markdown period to reduce the inventory and maximize margin over the whole period. This level of analysis can’t simply be carried out in Excel, by far the most common tool to plan markdown.

Over the years, many retailers have tried – and often failed – to introduce some systemization around pricing. However, the vast majority of those tools have been rules-based requiring the merchandiser to dictate the approach, which is no easy task. Fortunately, this is where machine learning is taking over. By employing advanced algorithms that can accurately evaluate price elasticities at this granular level, modelling millions of scenarios and evaluating the right actions for each item and store, the ideal price can be quickly and efficiently recommended.

And automated pricing does not negate the need for merchandisers. Setting the strategy is key, whether it be a focus on maximizing gross margin or establishing constraints around price setting. For example, many retailers are not wanting to have store-level pricing (unless at the end of markdown season), or perhaps prices need to be in round numbers or all items in a style set at the same price. Or perhaps there is a strict markdown budget. None of this negates the need to go granular, it is only by calculating at the lowest level and then optimizing across these constraints that the ideal price can be calculated.

It is only by introducing these tools that retailers can replace manual, simplistic, one-size-fits-all approaches to pricing with automated, strategy-driven intelligence free of human bias. This also helps retailers understand that markdown is not always a problem when managed well and integrated with the right buying strategy where risk is taken on the right items. It will mean discounts are applied appropriately avoiding where many items are today being reduced too much unnecessarily. Efficient pricing has the most direct impact on the profit and loss and for many retailers, AI-driven pricing has improved gross margin by 3-5% and bottom-line profit by over 10%. In these challenging times, achieving this dividend alongside the productivity improvements has never been more important. So please remember these simple rules:

  • Better buying reduces markdown but markdown is inevitable and desirable so your clearance strategy must be a critical part of your buying strategy.
  • Fixed markdown budgets control margin, not long-term profit, and a fixed depth of cut does not optimize profit, cash flow or waste.
  • Supply chain visibility is imperative to reduce markdown with too much safety stock costing money.
  • Understanding the right time to markdown is as important as how much to markdown and understanding where stock is located, not just how much, is imperative to optimize the price.
  • Wrong markdown pricing is hurting profits, and brand prestige; it is too important to be based on simple, rushed analysis and it is only by going granular and understanding the complex dynamics for item and store that pricing can be optimized.

So the next time you walk past a store with a sign saying “All items 50% off” you know why.