Applying the Trade Area Math to a Real East Bay Shopping Center

Published at May 5, 2026 ... views

#RealEstate#Retail#TradeArea#CaseStudy#BayArea#SiteSelection#OccupancyCost

The companion post in this pair walks through the mechanics of trade-area analysis — drive time, the 55–65% rule, demographics, leakage, category trajectories. That's the framework. This post is the case study that applies it.

Reading retail underwriting on its own can feel abstract: capture rates, sales-per-square-foot benchmarks, primary versus secondary trade areas. The numbers click only when you walk a real site through the sequence and watch the framework either confirm or kill the deal.

The trade-area framework gets tested once a developer puts it on a real site. The Fremont, East Bay site below shows the full underwriting flow — competition inventory, purchasing power estimation, supportable sales per square foot, and the rent number that decides whether the project pencils. Each step is a gate; if any of them fails, the project doesn't get built no matter how good the corner looks.

The case study runs about $113 million in addressable annual spending power, ~1 million square feet of competing retail at 89% occupancy, and a target sales-per-square-foot blend that has to clear a specific rent threshold. Whether that's enough is what the math actually answers.

The Fremont site, in context

Here's what this analysis looks like applied to a real project — a retail development site in Fremont, California, in the East Bay.

Fremont within the East Bay regional context, with the Ardenwood/Newark retail cluster pinned

The developer started with the basic site context. The site had good highway access and traffic, was near major employment, had planned new housing nearby, and faced limited retail competition. That's the high-level pitch — but the actual analysis is what determined whether to move forward.

Annotated aerial of the Ardenwood/Fremont site with I-880, employment nodes, and planned housing labeled

The Bay Area context is non-trivial here. CBRE's Q1 2025 Bay Area Retail Figures put the regional vacancy rate at 5.7% with average asking rent at $33.18 per square foot — a market that's tight on both fundamentals. Tighter regional fundamentals tend to give well-located new centers more room to lease up at acceptable rents, but they also raise the cost of land and construction, which has to be absorbed by the rent the trade area can actually support.

Street-level view of an Ardenwood-area arterial during weekday rush — the "traffic" half of the site pitch

Step 1: Map the competition

The first thing the developer did was inventory every shopping center within roughly 2.5 miles of the site. For each, they pulled the year built, square footage, occupancy rate, and anchor tenants.

2.5-mile trade-area radius drawn around the site with the 11 competing centers pinned

CenterDistance (mi)YearSq. Ft.OccupancyAnchor
Ardenwood Plaza0.8198832,000100%Round Table
Ardenwood Center0.8199238,000100%76 gas, Jack in the Box
Aspenwood Marketplace0.9200715,00057%Chipotle, credit union
Raley's Center1.01991120,00097%Raley's, Blockbuster
Newark Marketplace1.01993170,00081%Safeway, OSH, Starbucks
Lido Fair Center1.21980s100,00076%Ranch 99 Market
Rosemont Center1.21980s90,00061%Longs Drug
Northgate Center1.7197775,000100%Ranch 99 Market
Charter Center1.7198773,00092%Lucky Grocery
Brookvale Center2.41968131,00099%Lucky, Longs Drug
Comp set total / avg1,045,00089%

Total competitive inventory in the trade area: about 1 million square feet at an average occupancy rate of 89%. That tells the developer two things — the area can support meaningful retail (occupancy is healthy), and there's not a glut of empty space waiting to be filled (89% leaves room for a well-positioned new entrant).

Storefront of one of the named anchor centers — Newark Marketplace or Raley's Center — to put a face on the comp-set table

Step 2: Quantify the trade area's purchasing power

Next, the developer estimated total purchasing power available in the primary trade area, comparing 2008 (a baseline) to 2013 (current).

Choropleth of median household income across the Fremont/Newark/Union City census tracts

20082013Increase
Households5,5456,621+1,076
Avg HH income$121,118$127,300+$6,182
Total HH income (millions)$672$843+$171
Potential purchasing power (12.5% of income, $M)$84$105+$21
Total employees in area4,0004,400+400
Daytime convenience purchasing power ($M)$7.5$8.2+$0.7

The estimate uses about 12.5% of household income as available retail purchasing power for the primary trade area, plus a separate calculation for daytime employee spending — convenience purchases at lunch, after work, etc. Together: roughly $113 million of available retail spending power in the trade area, growing from $91.5 million five years earlier.

Loading diagram...

That number is the ceiling. The new center isn't going to capture 100% of it — existing centers are already capturing most of what's spent in the trade area today. The question is what residual share is available to a new, well-positioned entrant.

For a developer running this analysis on a different California submarket, two free data sources do most of the work the case study did manually. The CDTFA's quarterly Taxable Sales by County tables show actual retail sales activity by county and by type of business — the cleanest baseline for what's already being spent locally. Combined with population and household-income data from the census, you can reproduce the case study's purchasing-power calculation for any California submarket without buying a Claritas subscription.

Screenshot of the CDTFA Taxable Sales by County table for Alameda County — the free baseline for "what's already being spent"

Step 3: Estimate supportable sales per square foot

Finally, the developer pulled comparable sales-per-square-foot data for the categories they expected to lease to:

Wide interior shot of a busy weekend grocery checkout — the throughput that produces $/sf benchmarks

CategorySales / Sq. Ft. Range
Drug stores$513 – $797
Grocery stores$392 – $497
Restaurants$286 – $530
Services$143 – $266

Blending these by expected tenant mix, they landed on an expected average of about $450 per square foot for the subject property. That number then feeds back into the rent the developer can charge — if tenants average $450/sf in sales and want to keep their rent-to-sales ratio in the 5–10% range, they can support roughly $22–$45/sf in total occupancy cost.

Side-by-side storefronts of a drug store vs a service tenant — visualizing the 3.5×–5× spread in sales per square foot

That occupancy cost number is what tells the developer whether the project pencils. Run it through the cost calculator below to see how it shakes out:

📊

Total Occupancy Cost

What a tenant really pays to occupy office space.

Inputs
Results
Total Cost ($/sqft) $38
Annual Occupancy Cost $190,000

If the supportable rent number from the trade area analysis is at or above what the developer needs to hit their pro forma, the deal works. If it's below, the project doesn't pencil and either the cost structure has to change or the site has to be passed on.

That sequence — competition inventory, purchasing power, supportable sales, supportable rent — is the actual underwriting flow. Everything else in the project planning happens after these numbers come back positive.

What I keep coming back to

The reframe here is that retail underwriting is fundamentally a market analysis problem, not a design problem. The architect can make the building look great, but if the trade area doesn't support the rent, the great-looking building still ends up half-empty.

A newer suburban shopping center half-occupied, with one or two visible vacant bays — the "great-looking building, half-empty" outcome

That's why retail developers spend so much time on the front-end analysis — pulling comp sets, mapping demographic shifts, modeling capture rates, identifying leakage. The site doesn't get bought, and the building doesn't get drawn, until those numbers come back positive.

It's also why retail is one of the few product types where you can do a meaningful go/no-go decision off market data alone, before any design work happens. If the trade area can't support the rent, you can know that with reasonable certainty before you've spent a dollar on architecture.

I used to think site selection was mostly about picking a good corner with good visibility. The real work is much more analytical — and much more humbling, because the data sometimes tells you that the corner you wanted just isn't going to work no matter what you build there.

A few things I'm taking away from the case study

  • The actual underwriting flow goes: competition inventory → purchasing power estimate → supportable sales per square foot → supportable rent → does the project pencil. Skip any step and you're guessing.
  • Comp set occupancy in a trade area is a quick read on whether the area absorbs retail well. 89% across ~1 million square feet of competing inventory is a healthy signal; the same 89% across 100,000 square feet would be much less informative.
  • The 12.5% household-income capture rate plus separate daytime-employee spending is a usable rule of thumb for primary-trade-area purchasing power. Cross-check it against CDTFA county sales data when the project is in California.
  • The supportable rent number lives downstream of supportable sales-per-square-foot blended across the planned tenant mix. The 5–10% rent-to-sales ratio is the bridge between what tenants can afford and what the pro forma needs.
  • The interesting thing about the Fremont site isn't that the math worked — it's that the math could have worked or not, and the developer would have known either way before drawing a single building. Walk-aways are wins on retail sites.

This is the actual surface area of the trade-area framework: not just a circle on a map, but a sequence of gates that either pass or fail. Once they pass, the design follows. Once the design follows, the leasing follows. But it all starts with a careful, slightly uncomfortable look at what the people in the surrounding circle actually buy, where they buy it now, and whether there's enough of them to make a new center worth building at all.

A developer's site-visit kit on a clipboard — printed comp set, demographic map, calculator — the human-scale form of the analytical work

This post is the second of two on retail trade-area analysis in my ongoing series — Real Estate Development. The companion post covers the trade-area framework itself — drive time, capture rates, demographics, and leakage. Earlier in the retail arc: the percentage rent structure that defines retail leases and what shopping centers have to sell that Amazon can't.

Sources

Comments