Navigating Retail Vitality: Making Smarter Real Estate Location Decisions in a Changing Retail Landscape

We analyzed retail store openings and closings to identify where retail clusters (any area that has at least five chain stores grouped together) are growing or shrinking and categorize them by vitality.

• Out of nearly 50 thousand retail clusters, 45% have higher than average growth of number of stores, and 21% of all clusters have three times higher than average rate of store increases.
• Population growth is higher and unemployment lower in retail clusters with high and moderate openings.
• 37% of retail clusters have nearly the same number of stores after two years.
• 17% of all retail clusters have lost more than one store. The demographic most correlated to store closures is crime.

“Brick-and-mortar retail is dying” has long been a warning yet to be realized. In fact, many stores have shut down, but others have grown and thrived. Retailers want to identify either of those scenarios to locate in areas based on retail store count growth rate. “Retail vitality” indicates whether shopping centers are growing or shrinking to help retail real estate decision makers pursue a better location strategy and avoid costly mistakes.

When searching for new locations to expand their store footprint, retailers consider hundreds of demographics and market factors to identify which lead to higher or lower sales. One factor that’s top of mind for retail brands to understand is whether nearby stores are growing or shutting down. Examining whether stores are closing or opening can help identify the vitality of these dying or growing retail hotspots.¹

¹ To explore other themes driving retailer decisions, CBRE’s Five Forces Shaping the Future of Retail report dives into some of the underlying drivers shaping the trends revealed in this analysis.

CBRE Retail Analytics studied retail vitality shifts around the U.S. Using ChainXY data on store closures and openings of thousands of brands for the last two years, we built an algorithm that classifies every retail cluster into categories: High or Moderate Opening, High or Moderate Closures, Transitioning and Stable.

CBRE’s proprietary dataset, ShopoGraphics, generates a complete survey of the 50 thousand retail clusters that cover over 60% of all retail doors around the U.S. A cluster includes any area that has at least five chain stores grouped together. In each cluster we identify the number of stores that have opened or closed in the last two years. The following examples illustrate each classification type.

Figure 1: Cluster Classification to Measure Retail Vitality

High Opening:
Figure 2 shows a high-opening cluster around 8141 South Cicero Avenue, Chicago Illinois, surrounding a retail cluster we classify as Value Depot. This cluster is one of almost 1,900 around the country, in suburban or rural areas that usually contain auto parts, dollar stores, home improvement, lending and around 40 other value retailers.

The vitality analysis shows that in the last two years, 23 stores have opened and only eight closed. This cluster is classified as high opening.

Figure 2: A High-Opening Cluster

Moderate Opening:
Other clusters are growing their store counts, but more slowly. Those clusters are classified as moderate opening.

High Closure:
Compare that to another Value Depot cluster around 114 N. Vine Street, Urbana, Illinois (Figure 3). The vitality analysis shows in the last two years, seven stores have opened and 13 closed, including three big-box stores. This cluster is classified as a high-closure area.

Figure 3: A High-Closure Cluster

Moderate Closure:
If only a few stores close in the cluster as a percentage of total number of stores, the area is classified as moderate closure, to indicate areas that are shrinking but at a slower pace.

Transitioning:
This analysis identifies another area that ended up with the same number of stores with which they started. This Value Depot cluster around 1460-90 Douglas Road, Montgomery, Illinois, (Figure 4) has one fewer store than it did two years ago. But in that period, 12 stores closed and 11 opened. These zones are classified as transitioning.

Figure 4: Transitioning Cluster

Stable:
Other clusters have almost the same number of stores after two years but with no opening or closures. Those zones are classified as stable.

This analysis classifies nearly 50 thousand retail clusters into one of these types in to enable a consistent comparison of retail vitality anywhere in the U.S.

Benefits of Understanding Retail Vitality

Tactics will differ by brand: Some may want to find the growing retail areas and avoid the high-closure areas to position themselves as a premier brand. Others might want to take advantage of high-closure areas to expand their footprint at a reasonable cost. From a strategic perspective, brands need to know why retail is leaving some markets and growing in others. To understand this, we examined hundreds of demographic, co-locators and site characteristics of these retail clusters to identify the patterns that may lead to retail vitality.²

² For more research on what might be driving openings, closings, and opportunities, see CBRE’s Five Forces Shaping the Future of Retail report.

Growing Retail Clusters

Out of nearly 50 thousand retail clusters studied, 45% have higher-than-average growth of number of stores, and 21% of all clusters have three times higher than average rate of store increases.

Often the simplest explanation is the easiest, and store opening is correlated to population growth. Population growth is 1.1% higher in high-opening retail areas, and moderate-opening areas are growing at about average rate.

Also, the unemployment rate is 0.1% lower than average in high- and moderate-opening areas and 0.3% higher in high-closure clusters (Figure 5).

Figure 5: Store Opening vs Population Growth

Residential and workplace population is lower than average in high-opening clusters. This makes sense as these are growing areas, where the population density is expected to catch up over time.

One of the demographic characteristics that correlates most closely with retail vitality is crime (Figure 6). Both moderate-opening and high-opening clusters have below-average crime rates in every category we track.

Inversely, high-closures areas have higher crime rates.

Stable retail clusters have some of the lowest crime rates of all clusters.

Transitioning clusters have higher crime than opening clusters but lower than areas with moderate to high store closures.

Figure 6: Crime's Effects on Retail Vitality

Population growth and crime tell only part of the story. High- and moderate-growth areas had to be separated out because they looked substantially different, and we identified a few possible demographic and retail synergy differentiators.

High Openings:
High store openings often occurs in areas with lower household income. Median household income in high-growth areas is 1.8% lower than average, and median home values are 7.1% lower.

Retail clusters that are not in malls are more likely to be high growing. Big box/grocery centers, theme/festival/mixed/lifestyle and neighborhood centers are more likely to have high openings than other types, like large-format regional malls.

Moderate Openings:
Clusters with moderate store openings have high population growth and low crime, but some characteristics differ from high-opening retail clusters.

Clusters that grow around the average growth rate often include hardware stores, fast food, grocery, pharmacies and auto parts and repair stores.

Unlike high-opening clusters, these moderate-opening areas have a higher established residential and workplace population.

Power, regional, outlet or community shopping centers are more likely to have moderate growth than other centers.

Stable and Transitioning Retail Clusters

Around 37% of retail clusters have nearly the same number of stores after two years. Those can be separated into two distinct groups: stable and transitioning. The stable cluster includes the 15% of retail clusters that had almost no stores open or close. The transitioning cluster (22%) had nearly the same number of stores open and close.

Stable Clusters:
Stable clusters often have a high number of gas stations, hotels, auto rental and hardware stores. Areas with larger workplace population and those around strip centers, community and neighborhood malls are less likely to be stable.

Stable clusters, like moderate-opening clusters, have relatively low crime compared to other vitality categories.

Transitioning:
Transitioning clusters usually have an above-average number of discount and big-box stores in areas with a higher population. Power and regional centers are more likely to be transitioning, along with 46% of large-format regional malls.

Transitioning areas have higher-than-average median household income.

Shrinking Retail Clusters

In contrast to growing or stable clusters are those that have more store closures than openings. Around 17% of all retail clusters have lost more than one store, including both moderate closures (10%) and high closures (7%).

Store closures correlate closely to crime. High-closure retail clusters on average have 2.13 burglaries per 1,000 people, while high-opening clusters have only 1.75.

Figure 7: Likelihood of Store Openings by Geographic Type

Moderate and high closures are more likely to happen in urban and hyper-urban settings.

High-opening clusters are more likely in rural or fringe areas where the population is growing and retail is expanding.

Transient and workplace population percentages are higher around high-closure areas than high-opening areas, which may indicate challenges for urban retail. Also, higher home values correlate to store closures, another tailwind for urban retail.

Growth and closure rates, and thus retail vitality, also vary by shopping center type. Strip and neighborhood centers are much more likely to be in the high-openings cluster. Power centers appear least likely in high-closure clusters. Super-regional malls have the highest proportion of closures and transitional clusters, where almost as many stores opened as closed.

Figure 8: Store Openings by Retail Format

This data gives an objective, systematic method to examine how retail is evolving around the country. We examined hundreds of other demographics and co-locators datapoints to see what patterns may be causing some of these retail vitality changes. We entered hundreds of datapoints into a machine learning algorithm to predict which cluster may be opening or closing more stores in the future.

Knowing if a retail cluster is growing or shrinking and understanding the possible causes of retail vitality are useful, but predicting what will happen to these clusters in the future is the goal. Machine learning can be used to identify Opening, Transitioning or Closing clusters. Our machine learning algorithm accurately places most shopping center categories where expected and highlights some retail clusters that are likely to start growing, shrinking or changing due to market conditions (Figure 9).

Figure 9: Store Openings – Actual vs Predicted

Brands can elevate their real estate strategy using data-driven insights to avoid struggling retail clusters and take advantage of the growing retail hotspots. These insights can be used to understand how these vitality trends will shape the future of retail.

Figure 10: Map