The global luxury goods market has enjoyed spectacular growth over the past 30 years. We predict this will continue based on an analysis of luxury goods retail stores. We mapped Burberry, Chanel, Gucci, Hermes, Louis Vuitton, Prada, and Apple stores around the world by city. In total 2,176 luxury good / high fashion stores.
Below is a map with all the stores.* It’s obvious that large, affluent cities have many stores. There are also many vacation destinations with stores.
CLICK ON THE MAP TO ENLARGE IT
What is less obvious from a visual inspection is:
- How does income influence store density?
- Are there other factors than income that explain store footprints?
- Which cities are over- or underpenetrated?
- Where is the market heading (is there a plateauing effect)?
We tested this with a nonlinear regression using a Gompertz function** on the city dataset:
Number of stores = f [ Population with income over a certain income in the city, Capital dummy (city is capital = 1, not a capital = 0) ]
The results are plotted with the actual and predicted number of stores on the y-axis, and the upper middle and upper classes on the x-axis (logged). The fit is good.
This seemingly simple analysis tells us:
- The best explainer of the number of stores is a city’s upper middle class and above population (not lower middle class and above; not middle class and above; not upper class; but exactly upper middle class and above = AB and C+ socioeconomic classes)
- Being a capital has a strong positive effect on number of stores
- There is no plateauing effect. Most categories exhibit an s-curve when we do this type of analysis. There is no evidence of such an s-curve yet, Instead, the potential is infinite.*** We see exactly the same in other premium/superpremium categories.
- Beyond this, certain countries have a “love of luxury”
Using this knowledge, the following cities are over- or under-penetrated:
And we can project the number of stores 10 years from now:
To learn more, please contact one of the co-leaders of our luxury goods practice:
Or call the number at the upper right.
— — — — —
* Includes company-branded stores; excludes department stores, stores-within-stores, etc.
** The Gompertz function will identify s-curves, if they exist.
*** This will change sometime in an unknown future, but for the time being there are no signs of limits / plateauing.