Figure 1 illustrates the trade area performance by level of competition for each retail chain. It splits each retail chain into three parts: the portion of their network in high, medium, and low levels of competition. It reveals a surprisingly large retail cluster trap across virtually all retail categories as revealed by a strong correlation between high competition levels (red dots) and poor performing trade areas (bottom right quadrant).
Over-clustering effectively creates something called Nash Equilibrium: which is typically results in a high cost, low revenue retail trap where retailers have no incentive to relocate. Relocating means relinquishing your clientele to the competitors across the street. However, not relocating means an over-competitive no-win market environment that limps along with lower profits that limit the ability to expand to better locations over the long run. In the end, the cluster trap leaves retailers with the only practical option of spending more on marketing, store experience and customer service, or to lower prices, to try and increase sales.
Over-clustering creates hit-and-miss networks where there are too many retailers in some areas and not enough in others. We evaluated 65 retail categories in Canada and found that 53% of retail suffers in low revenue, over competitive retail clusters. And yet 35% of trade areas have no competitors at all and half of those are excellent opportunities. This represents an enormous opportunity for business expansion and contraction in every retail category in Canada, including those in the best performing categories.
Figure 1: The Retail Cluster Trap using STRATA Trade Areas
How Retail Cluster Traps Occur
There are two major reasons the cause the retail cluster trap. First, current location methods are flawed. The problem is how consumer behavior is modelled. Current location methods limit consumers to buying goods in one trade area at a time. They can’t account for the real-life complexity of consumers buying across multiple trade areas.
The problem stems from the fact that trade areas need to be large enough to accommodate the complexity of consumer choice. Trade area leakages occur when consumers purchase goods and services in both their home and adjacent competing trade areas. This means that when you map out customer sales, you end up with trade areas that overlap adjacent ones. This means that a simple trade area definition for a shopping centre includes consumers from other competing shopping centres. As a result, circles or drive-time trade areas, and mobile data with geo-fenced derived trade areas double-count market size often by the number and amount of overlapping trade areas. This exaggerates the anticipated profitability of a trade area and therefore the number of competitors a trade area can support. This in turn creates the retail cluster trap marked by high cost, low revenue networks.
Figure 2 analyzes the performance of Canadian retail chains using 3 km circle trade areas by level of competition. The quadrant analysis shows that there is a strong relationship between high revenue potential trade areas and high competition. It says that highly clustered trade areas are also the highest revenue. This suggests that Canadian retail networks have primarily been built using simple trade area definitions that can't account for consumers buying across multiple trade areas.
Figure 2: The Retail Cluster Trap using 3 km Circle Trade Areas
Figure 3 shows a performance quadrant graphic for 13 retail product groups. It shows that Food & Drink, Recreation, Restaurants and Personal product groups are some of the best performing retail groups. Best performing means that good performing trade areas outnumber poor performing ones. Interestingly, these groups also have some of the lowest levels of retail clustering and some of the best feedback loops of all retail categories due to the frequency of customer shopping.
Retail categories with a high frequency of visits have a much quicker feedback loop which often leads to overall better performing networks. In contrast, retail categories like Vehicles and Financial related categories have limited feedback after the initial visit and hence have poor performing networks. This reiterates that categories without a strong feedback loop tend to have weaker networks because they lack consistent customer visits which provide a measure of network performance. It also makes low-feedback networks the most precarious because they rely on so few consumer visits. This puts an enormous amount of importance on location choices, especially for retailers with few physical customer visits. When a customer visits a retailer only once a year, such as a bank, wealth manager, insurance broker or tax consultant, the location becomes exponentially more important to the success of their business.
Figure 3: Store Performance by Retail Product Groups