We stopped drawing ring studies for internal use in 2019, and we have not put one back into an investment committee memo since. This is a note on what replaced them, why it matters, and what it actually changes in how we price a center.
The ring problem
A concentric ring study asks the wrong question. It asks who lives near this center when the real question is who shops at this center. These turn out to be very different populations, and the gap between them is where most underwriting errors hide.
There are three specific failure modes worth naming.
Highways and drive-time moats. A ring treats a half-mile strip of interstate as if it were a neighborhood. Customers do not cross it. We own properties where the eastern half of the three-mile ring contributes less than 8% of the center's actual visits, because a four-lane arterial and a canal make it a twelve-minute drive instead of a four-minute drive. Under a ring, those households are customers. Under the mobility data, they do not exist.
Asymmetric catchment. Real trade areas are lobed and biased along the direction of daily commuting patterns. A center on the west side of a major employer will draw from the north and south and almost nothing directly from the east, because nobody commutes outward in the morning to then shop on the way home. A ring smears this into a uniform circle and calls it a demographic profile.
Competitive overlap. A ring counts every household inside the radius as a potential customer, but a household with a closer or more convenient competitor is not, in any meaningful sense, a customer. The correction for this is not a ring. The correction is a probability model.
A ring study tells you what the neighborhood looks like. It does not tell you what the business is.
Drawing a trade area honestly
The modern way to draw a trade area is from mobility data. A mobility dataset — we buy from one of the two or three credible providers in the category and layer our own cleaning on top — takes anonymized device pings from phones that visited a given location over a rolling twelve-month window and tells you, probabilistically, where those devices spent their overnight hours. In other words: it tells you where the customers actually live.
From that, you build a visit-weighted trade area. Not a ring. A shape. It typically looks like a lopsided ellipse stretched along the direction of the primary arterial, with chunks missing where competitors capture the overflow. That shape is the honest answer to "who shops here," and it is the starting point for every other piece of the underwriting.
The analysis then extends one level deeper. For each device in the trade area, you can look at how often it visits and how long it stays. A household that visits weekly for 22 minutes is a fundamentally different customer than one that visits monthly for 6 minutes. A tenant mix built on the first is much harder to disrupt than one built on the second, and a center's durability against a new competitor entering the trade area is almost entirely a function of how sticky the existing visit cadence is.
A case, anonymized
Consider a center in our portfolio that was underwritten in the second half of 2023. The three-mile ring showed a population of 64,000 with a median household income of roughly $58,000 — a perfectly ordinary working-class Florida trade area.
The mobility-derived trade area showed something else. Population 38,000, median household income $71,000, concentrated heavily to the west and north, with the eastern half of the ring essentially absent from the customer file. Same physical building, same GLA, and two entirely different retail businesses.
Under the ring, the property looked like a value acquisition in a middling trade area. Under the mobility model, it looked like a slightly smaller but substantially wealthier catchment with tenant demand we could price differently. The second analysis was the one the leasing outcomes over the following eighteen months validated.
What it changes in rent
Three things change, concretely, when you make the switch.
- Rent projections get tied to the visiting population, not the resident population. That tightens the inline rent band, in our experience, by 40 to 50% on the downside and opens meaningful room on the upside for well-positioned trade areas.
- Tenant curation is built around the demonstrated trip purpose of the actual customer — not the theoretical demographic. We have declined tenants that the ring study said were a fit and the mobility data said were already overserved elsewhere in the catchment.
- Competitive exposure becomes auditable. We can see, in advance, where a competing grocer or a new power-center development is likely to bleed us, and we model the probability of it before we sign anything.
The cost of still using rings
Our working assumption is that roughly half of the private owners bidding on community retail in Florida are still underwriting with ring studies. We are comfortable with that, because it means their acquisition math is systematically wrong in the same direction — and we occasionally get to buy centers from them at a price our model says is generous and theirs says is a mistake.
The tooling to do this properly is no longer expensive. The reason most small private owners do not use it is not cost. It is that they have not updated the procedures they learned in 2006, and the procedures from 2006 are the reason 2006 acquisitions looked better than they turned out to be.