This article is adapted from “Boosting mall revenues through advanced analytics,” first published in January 2019 on McKinsey.com.
With consumers shifting more of their spending from physical stores to e-commerce and increasingly looking for experiences rather than products, mega-mall operators worldwide are seeking ways to reinvent their business. Today’s most successful megacomplexes aren’t just sprawling collections of stores but rather carefully planned “retail-tainment” centers—destinations for shopping, dining, entertainment, and a wide range of other leisure activities.
Smart choices about tenant mix and placement, space allocation, and rental values are critical to the success of such large-scale retail developments. In this article, we home in on how retail developers and megacomplex owners are leveraging advanced analytics to make crucial business decisions. Specifically, they’re using advanced analytics to plan store adjacencies that yield higher consumer spending and longer mall visits, and to engage in more-informed rent negotiations with tenants. It’s paying off: large-scale retail developments that use these tools have increased their revenues by double-digit percentages.
Matching tenants to locations
When a leasing team reviews leases set to expire in the next quarter or year, it should study the universe of potential tenants to fill the pipeline: current tenants that might be better off occupying a different unit within the mall, tenants that are in the company’s other malls but not in this one, and any potential new tenants that have expressed interest in leasing a unit.
Mall owners sometimes remodel select “precincts,” or sections of a mall devoted exclusively to certain product categories. These remodeling projects provide an opportunity to move tenants to different units within the mall. Assessing all the possible combinations and permutations is a complex exercise that requires the power of advanced analytics.
New tools—let’s call them “right tenant, right location” (RTRL) tools—can estimate each tenant’s omnichannel sales and the rent it should pay, in every potential combination of tenants and units. (Note that in determining tenants’ rent, a mall operator’s primary goal shouldn’t be to maximize its own leasing revenue but rather to maximize the combined sales of the mall’s tenants. That may seem counterintuitive, since tenants’ sales don’t directly translate into bottom-line impact for the mall operator. But as each tenant’s occupancy cost ratio2 goes down, the mall’s tenancies become healthier and more sustainable, helping the mall remain viable.)
The aforementioned company used an RTRL tool to figure out which ten of its prospective tenants should replace the ten current tenants whose leases were expiring, and which specific units in the mall the new tenants should occupy. The tool showed that, with the new set of ten tenants, the mall could generate an additional $115 million in sales compared with the prior year, plus an additional $5 million in leasing revenue (see exhibit).
Sales and productivity data make up the backbone of an RTRL tool; its predictions will be only as good as the information it has on both current and prospective tenants. The mall owner should gather enough reliable data—ideally, not just on omnichannel sales but also on occupancy costs—to enable the tool to differentiate a tenant’s potential performance in unit A compared with unit B. If the mall owner has no data on a prospective tenant, it should carefully select benchmarks from its current portfolio of stores.
An RTRL tool must also be able to take business constraints into account. For instance, if a mall has precincts, the tool shouldn’t assign a toy retailer to an available unit in the apparel precinct or the food court. In addition, the mall owner should specify a minimum or maximum category share—for example, high-end apparel at 10 percent of a mall’s total square footage, low-priced electronics at 3 percent, and so on.3 Without sensible business constraints, the tool will inevitably recommend placing high-productivity categories (such as luxury accessories or consumer electronics) in unreasonably large stores.
Setting a price range for every lease
One major challenge for mall owners around the world is determining rent for each tenant. A typical mall doesn’t have clear processes or a shared understanding within the organization on how to set rental targets for each tenant; the asset-management and leasing teams, for example, often arrive at different targets. Publicly available industry benchmarks, by country or category, don’t exist. Our own research suggests that a mall’s leasing revenue expressed as a percentage of total tenant sales varies widely—from 5 percent for a mall in Brazil to 25 percent for one in Australia.
In theory, the rent for a unit in a mall depends on four variables: the type and location of the mall, the quantity and quality of foot traffic, the characteristics (including size and configuration) of the unit, and the sales productivity of the brand or category that will occupy the unit. In practice, however, it’s a different story. We’ve found that leasing managers base a unit’s rent on personal knowledge of the tenant (“I know him, he won’t pay more than $x”), past practices (“Let’s increase the rent by 2 percent this year, like we’ve done every year”), or a combination of gut feeling and business acumen.
A clear view of the overall and unit-level economics of the tenants, combined with an advanced-analytics pricing tool, helps a mall owner in Asia to set better rental targets. The company conducts a thorough analysis of the economics of each of its tenants, studying not just sales per unit but also occupancy costs and profits. It then uses a pricing tool that breaks down the four variables previously listed into approximately 25 subvariables, such as the unit’s proximity to a mall entrance and the brand’s price positioning. Using multivariable regression, the tool sets a price range (or a “zone of possible agreement”) for every current or prospective tenant. Calculating value creation at the unit level—and coming up with a unit’s estimated rental value (ERV) for each tenant—thus becomes a much more rigorous process. Use of the ERV tool has put the mall operator on track to capture a 20 percent increase in leasing revenue over five years.
Advanced analytics equips a company to manage and improve not just its short-term performance but also its long-term health. One company’s ERV tool made clear that the mall’s low-performing tenants were paying rents that were disproportionately higher than the mall average—effectively subsidizing other tenants. This was an unsustainable situation: sooner or later, the overpaying tenants would present a considerable vacancy risk, particularly if they were concentrated within a single leasing cycle or mall precinct. The tool also showed that the mall owner could negotiate more assertively with its higher-performing tenants. In such situations, implementing the tool and acting upon these types of insights can yield up to 5 percent of earnings before interest, taxes, depreciation, and amortization per year.
For a case example of how advanced analytics can help a mall quantify each tenant’s sales impact on other tenants, read the full article, “Boosting mall revenues through advanced analytics,” on McKinsey.com.