Queues are bad. Really bad. They literally cost you money. One of the principles of queue theory is that once a queue forms it has to be managed, which requires resources, then you need over capacity at some point, in order to reduce the length of the queue. Often the resource to manage the queue comes from the very thing serving the queue. This means the queue gets processed even more slowly than before the queue started and the queue gets longer. It doesn't start to get shorter until you have over capacity.
Retail queues cost money in all kinds of ways and often those costs are hidden, or at least not immediately apparent. Long queues lead to customer dissatisfaction and ultimately to lost customers. Losing customers obviously isn't a good thing and will cost you money in the long run, and it's in this respect that most people associate the cost of queueing. (Queueing, incidentally, is the only word in the English language to contain 5 consecutive vowels)
Some less obvious ways that queues cost money involve
- In a supermarket, the space kept clear between the checkouts and the start of the aisles is valuable space that is being used to manage queues instead of being used as shelf space.
- Two delivery lorries arriving at a store/depot at the same time cost money in waiting time.
- Or a more trivial example, taking a ticket at the meat counter involves the cost of the little yellow tickets and the dispenser
Understandably therefore, minimising queues is top priority, and as something that has big impact on customer dissatisfaction, it's something of a PR tool in the supermarket wars. The trade off is balancing the cost of resources used to prevent queues, vs the cost of the queue itself. Having all checkouts open all the time would go a long way to easing congestion in the store, but that has to weighed up against the retail wage budget, (the single biggest expense for a large retailer), used to pay all of those checkout operators required to ensure no queues.
Retailers understand this trade off and have developed complex systems to help them manage queues, something that the average shopper trying gauge whether aisle 19 (short queue, but inefficient operator) is better than aisle 20 (longer queue but smaller basket size) probably doesn't appreciate. (Fuzzy logic might help here!) Space in store is critical and the retail wage budget is massive, so what looks like large investment can be easily offset by the cost of real savings in either of those areas.
You might like to ponder this next time your stuck in a checkout queue, It's something I can't help myself doing... (My personal favourite is the Wilkinson's dot matrix receipt printer, which takes something like 15 seconds to print your receipt. Whilst laser printer are more expensive, the retail wage budget saving, at roughly a quarter to half a million pounds per second, I'm sure would massively outweigh the investment)
You can apply a queue theory equation in loads of different ways (another favourite of mine is NHS waiting lists). In a later post I'll talk about how I've used queue theory to revamp a large data warehouse batch schedule..