00:31 What is so interesting about the assortment which is offered in a luxury store?
01:34 What sort of characteristics do hard luxury stores have? Where are they located? What is their target market?
03:04 What are the key challenges hard luxury stores face?
05:08 With such a sparse amount of data, do we have enough to infer valid statistical conclusions?
08:34 By having that understanding of what individual clients want, you can ensure that most likely in two or three years’ time, when they will make their next purchase, you will have the next big thing ready in store for them. Is that right?
11:25 The main aim of the game is to have the optimum assortment for each store, for all the clients who frequent that store. How easy is to do this in reality?
14:38 Let’s imagine I am a brand with 20 units to sell and I need to know where I should place them and when. What do I actually do?
19:51 What are the first steps you should take to implement an optimization model?
The nature of the hard luxury market means that despite a growing online presence, the vast majority of purchases are still made in store. With sales reaching as low as one or two units per year, it can often be challenging to know what to display in stores and how best to capitalize upon a demand that is so sporadic. For this episode of LokadTV, we discuss this dilemma and learn what hard luxury stores can do to optimize their assortment.
What is very interesting in hard luxury is that it’s the stock itself that creates demand. It’s somewhat rare for somebody to buy a very expensive item without seeing it “in the flesh”, so it’s usually the experience in the store that leads to a successful sale. Hard luxury, much like fashion, is driven to a certain extent by novelty. One of its major characteristics is that there is very little data, due to the highly sparse sales. In addition, the classic time series forecast, that is ubiquitous in supply chain science, doesn’t work at all for hard luxury. However, just because time series may not work this doesn’t mean that the only way to go forward is with pure “guesstimation”.
What needs to be done is considering the network as a whole in order to gain other perspectives. For example, the use of customer loyalty cards and other customer data to be able to link customers to products. This way, assortments can be picked and optimized in relation to the clients that are likely to visit that store. It’s certainly not easy to do, but not by any means impossible. For a start, hard luxury products have excellent traceability, which helps greatly.
To conclude, we talk about the use of probabilistic forecasting to assess the performance of assortments, the possible cannibalizations that can occur and to make sure that they get distributed to the correct stores and not just the best performing stores. No matter what, there is an irreducible uncertainty in hard luxury, but probabilistic forecasts allow much clearer and more powerful assessments.