Are all UK shopping centres the same?
A data led analysis of the biggest centres and their shops
In the mid-80s to early 90s going to shopping centres was becoming a rather homologous experience. You knew what to expect; it didn’t seem to matter where you went they all looked the same, seemed to be called the same, and have exactly the same shops. At the same time they just seemed to get bigger and bigger; culminating in the opening of the biggest of them all, the Metrocentre in Gateshead, in 1986.
Now we are repeatedly told by the press that retail is dying and that there are too many shopping centres (currently around 550) in the UK. Across the country, BHS premises remain empty, and even Marks & Spencers, the cornerstone of UK retailing for so long, has announced closures.
Online shopping is killing retail. ASOS which has exactly zero shops is now worth more than Marks & Spencers, which has 900+ stores. Living standards are being squeezed by austerity measures and everyone seems to be looking for a bargain. Moreover social media means teenagers have don’t have to hang out at shopping centre to talk to one another. It’s worse in America, where malls defined a whole generation and culture, it’s predicted that one in 4 US malls will be shut in 2022. (Time Magazine, Why the Death of Malls Is About More Than Shopping?, June 2017)
Over Christmas we found ourselves for the first time in ages in shopping centres with our families. What struck us was that we didn’t recognise half the shops. Perhaps we’re getting old, admittedly we were still looking for Woolworths and Our Price, but something seemed different.
So a question was formed.
Do all shopping centres still have same shops?
Again a simple question leads to a long time collecting and checking the data (there appears lots of money to be made selling retail data, but not much seems free). We scraped retail listings from some of the top UK shopping centres and, after much manual matching of retailer names, were able to pull together a dataset to help analyse what shopping centres look like today.
What is interesting to note initially is the size of different shopping centres across the country. The top 5 occupying a league of their own compared to the rest.
If all shopping centres had similar shops we’d expect a large number of the shops in each centre to be in other centres. This is certainly the case for some; EE, Debenhams, Claire’s,The Perfume Shop, and the Carphone Warehouse (offering an interesting mixture of different retail types) have the highest number of stores in the shopping centres that we have data for. Although no retail chain had a shop in every shopping centre.
However looking across the shopping centres, ordered by their retail size in square metres, it is surprising how many centres have a high proportion of shops which only feature in less than five other shopping centres. The big chains, with more than 20 stores across the UK, make up between 20 to 40 per cent of the shopping centres.
Looking geographically next we wanted to map out if there was a difference between areas across the country in reference to the unique stores. So we built a map, a strange map, but a map nonetheless. Each shopping centre is positioned in its rough geographical location in the UK (so Metrocentre is north and Bluewater is south) and each square then represents a store within the shopping centre. The single stores, stores in only one location, are then highlighted.
And there seem to be a lot of single stores, the big centres of Metrocentre, Bluewater and the two Westfield’s have a larger proportion of single stores. One theory could be that centre managers are filling vacant space by offering cheap rent to entice new tenants. Or that the stores in these bigger centres attract more unique, destination, shops for online only services.
Note that on our data Eldon Square and Victorian Centre have no unique shops.
What about the type of shops? Shopping centres seem to have a wide range of different types. Fashion and food and drink dominates in the majority, but each of the ten segments we have identified are well represented. Clearly shoppers demand a decent brand mix and leisure offer.
Our analysis of the UK’s shopping centres shows different shopping centres do give a different ‘retail experience’. Whether this has always been the case is difficult to know as we don’t have access to historical data but buying online has been a retail game changer for shopping centres. In the past stores needed about 250 stores to the cover the country, but now with an online presence, stores only have to open a much smaller number.
One future for shopping centres could be a division between those with high-end destination shops, where bagging a bargain is not a primary concern for customers, compared to centres full of budget stores, where bagging a bargain is definitely the primary concern.
So maybe it isn’t all bad news, as the future might offer a valid way for them to compete and coexist with Web 3.0. Viewing and trying high-end products we buy online will be needed until virtual reality is good enough to help, and people may never substitute VR for the real thing. Shopping centres could, one day, become no more than huge window displays for the products they stock, with deliveries automated to your home on ordering. A fleet of drones could then be delivering your shopping and have it waiting for you when you arrive home. Scary….
All we would need to answer the question was get the store data from a number of top shopping centres across the country and then compare. Simple.
Three problems. One, scrapping the different centre websites, which are mostly awful, isn’t fun (we used Parse Hub https://www.parsehub.com a free web scraping tool that is relativley easy to use, although at times it seemed to have a mind of its own).
Two, no shopping centre can agree what a shop is called. Is it Three or 3 or Three Store or 3 (Three) Store. No-one seems to know. So we had to manually check the store names from different centres, and then make a judgement (or guess) if the they were indeed the same. This ended up being for over 5,000 stores from 37 of the UKs biggest shopping centres.
Three, the type of shop required a very subjective labelling, mostly by hand, as most of shopping centre websites coded stores differently. It usually required a quick google check to find out what the store actually sold. Most of the time we didn’t have a clue.
Our data cannot then be considered 100% accurate.
Tableau link below :
And data :