Using a Capacity Rationing Strategy for Optimising Sales within a Dual Serving System

Online daily deals platforms have been expanding greatly in the online retail sector. The launch of Groupon in 2008 led to a 332% annual average growth over five years in the US daily deals industry. This impact is expected to grow globally, as, today, even big e-commerce platforms such as Amazon and eBay, whose fundamental aim is to shift trade transactions from offline to online, are offering daily deals.

When browsing a deals platform online, we might question how a product or service can be cheaper here, than on the original supplier’s website?

There are several answers, e.g., the dealers buy in bulk with good quantity discounts, so can accommodate their margin, and/or, they have very little operational and logistics costs, so can afford a very low margin, and/or, the supplier wants to introduce the product and penetrate the market, so, bears low/no margin sales for a batch of products, etc.  

However, a more difficult question is whether and how the supplier should make a deal with these platforms, which we call ‘merchants’. On the one hand, the supplier is (usually) offered a large purchasing order, which is great, as it sells the stock much quicker than the regular sales do, and on the other hand, a higher profit margin can be obtained if the supplier waits for the regular customers.

You might ask why not plan for sufficient capacity to serve both
demands (from merchants and regulars)? A big issue is that the deal offers are highly unpredictable and unstable (both in size and occurrence), thus, if these are considered as the actual demand and aggregated with the regular demand, they would cause a big amount of variability, which would lead to a high ‘buffer’ capacity (of stock or service) that is costly to maintain.

We call this situation the “dual serving problem”, where, a supplier serves its regular customers as well as merchants who sporadically
show up with large orders. Each of these groups of customers make their purchases according to their needs, resulting in two different patterns of demand. The regular customers’ demand is very frequent and stable, but for a small/medium amount, while the merchants' demand is very occasional, but in very large quantities.

To cope with the problem, we propose a ‘capacity rationing strategy’ that works as follows: any demand from the two groups (of customers) is fulfilled immediately on a first-come-first-served basis, as long as the available stock is above a certain ‘protection level’, under which only regular customers are served. A new replenishment order is placed as soon as stock on hand drops to the protection level.

We defined the optimal settings of such a strategy using a mathematical model (with some assumptions) and analysed it numerically via a simulation model. The findings prove that, in certain circumstances, the protection-level strategy significantly outperforms the traditional (demand) aggregation approach, and has a great potential for successful implementation wherever the dual serving problem exists.