Free Subscription

  • Access 15 free news articles each month


Try one month for $5
  • Unlimited access to news,insights and opinions
  • Quarterly and weekly magazines
  • Independent research reports and forecasts
  • Quarterly webinars with industry experts
  • Q&A with retail leaders
  • Career advice
  • Exclusive Masterclass access. Part of Retail Week 2021

Allocations – black art or science?

clothes, fashion, shoppinng, womenswearWe’ve all read it a million times: right product, right price, right place and right time. These are the four critical things retailers must turn their minds to with every decision, yet strangely most retailers pay lip-service to the ‘right place’ factor when it comes time to allocating stock.

Even in this era of fairly sophisticated allocation algorithms, I suspect that a lot of black art is going on as the allocators try to manipulate the computer-driven result to impart their “special” knowledge.

Before we examine the logic of the algorithm, it is worthwhile stepping back to the ordering stage. Fashion retailers have a process called “sizing” whereby they determine the optimal size curve based on a like-style’s sales history. It’s not that difficult to work out the ratio across the size curve but there will be some compromises needed, for example, in the table below Size 8 is ordered at 10 per cent but the like-style sales history is only 7 per cent, in other words, it is marginally over-bought. On the other hand, sizes 10, 12 and 18 are marginally under-bought.


If the buyer has determined that we will buy 1,000 units, then we order 100 packs of 10 in the example. Sizing is a deceptively easy process.

Roll the clock forward a couple of months and now it’s time to allocate this order. Let’s assume that we have 80 stores for this stock. One scenario could be one pack for all the stores and an extra pack to the largest 20 stores. What happens if the largest store’s turnover was five times higher than the smallest store? The largest store is massively under-allocated as it only received twice as many packs as the smallest. The problem is not really capable of being solved at this stage in the supply-chain. The solution really had to be hatched back at the ordering stage where perhaps some further compromises were needed:

  • create smaller packs – which deviates from the ideal size curve
  • a combination of small and larger packs – which might leave some stores with a less than ideal size curve
  • more stock in total – which might not be justified according to the buyer’s view
  • tweaking the assortment plans (that is, not an all-store buy) – which will leave some stores with no stock.

Sizing is an absolute misnomer. It should be called pre-allocating and it is an extremely important function. Yes, sizing is about the right product but it is inextricably linked to the right place.

No one is in a position to critique all the allocation systems in the market, however, it is possible to set down some underlying principles that ideally should be reflected in the algorithms.

Rule 1.  Allocate to where the warm bodies are shopping

Sales are clearly always going to be the main driver in any allocation. But when we talk about sales we really mean a mixture of what’s happening now and what we think is about to happen based on history. This is the so-called “sales need”.

Rule 2.  Allocated to stores that are selling it fast

Stock drives sales. A fast stock turn is great but it represents lost sales opportunities, so it is a valuable piece of information that should be factored into the allocation algorithm. Many algorithms simplistically determine a “stock need” based on re-filling stores to a pre-determined stock level regardless of sales potential. They then add “sales need” to “stock need” to come up with a “total need” and allocate as closely as possible to that final number.

The problem with that method is that stock need should be calculated on relative speed. If a store is selling out twice as fast as the company average, give it twice as much stock. By doing that, lost sales opportunities are inherently captured in the algorithm with a much better result.

In conclusion, don’t dismiss sizing and allocating as simple exercises. They can be deceptively complex tasks. Firstly make sure your allocation system factors in sales and rate of sale not sales and stock. Secondly, ensure each and every order is given due attention early in the buying and supply chain process before it actually comes time to execute the allocation – by then it’s too late.

Graham Lack has over 35 years retail experience in senior management roles at Luxottica and Suzanne Grae, in retail operations, finance, IT, marketing, merchandise planning and logistics. Contact him via

You have 7 free articles.