Demand Analysis in Feasibility Studies: How AADT, Capture, and Mobility Data Drive the Forecast

How institutional demand analysis is constructed for traffic-driven retail under SBA SOP 50 10 8 and USDA 7 CFR 5001, illustrated through gas stations, express car washes, and truck stops.

The number that everything else hinges on

In every feasibility study, one number propagates through the entire model. Capture rate. Whether the project services its debt depends on whether the analyst can defend that single figure to an SBA underwriter or USDA reviewer. Demand analysis is the work that gets the analyst there honestly, or fails to.

Most studies that come back for revision do not fail on math. They fail on the demand chapter. The trade area was defined by radius when corridor analysis was appropriate. The traffic count was used gross when only one direction matters. The pool of demand was multiplied by an industry rule-of-thumb capture rate without first netting out the four competitors already operating inside the trade area. By the time the pro forma is built, every line item carries the error.

The institutional method is different. It defines the trade area first, then quantifies the demand pool inside it, then nets competitive supply, and only then derives a capture rate from what remains. Each step is documented. Each step is triangulated against an independent source. The output is a number a lender can underwrite, not a number an analyst can hope for.

Demand analysis in the broader feasibility framework

The Appraisal Institute codified the structure in The Appraisal of Real Estate, 15th edition and in Stephen Fanning's Market Analysis for Real Estate. Six steps in sequence. Property productivity, market area definition, demand analysis, competitive supply analysis, residual demand calculation, subject capture forecast. The demand chapter is Step 3, and it has a specific job. It quantifies the units of demand the trade area produces, expressed in whatever metric the asset class actually consumes. Gallons for a gas station. Washes for a car wash. Truck fuelings and parking events for a travel center.

Four concepts are routinely conflated and should not be. Market sizing answers how big the universe is. Trade area analysis answers where customers come from. Demand analysis answers how many demand units exist inside that trade area, and what share the subject can realistically capture. Competitive analysis answers who else is fighting for the same demand. The four are sequential, and the work upstream determines the integrity of the work downstream.

SBA SOP 50 10 8, effective June 1, 2025, sets the regulatory floor. The SOP keeps gas stations, car washes, and truck stops on the special-purpose property list, which triggers heightened appraisal and feasibility scrutiny. Appendix 7 imposes Phase I ESA on every gas station loan regardless of size, and the recently expanded NAICS 484 trucking listing pulls truck stops into the same environmental scope. Where the lender's own analysis cannot establish reasonable assurance of repayment, an independent third-party feasibility study is required. Demand analysis is the analytical core of that study.

USDA OneRD applies a parallel framework under 7 CFR Part 5001, Appendix A to Subpart D. The market-feasibility lens explicitly requires the consultant to address demand, competition, pricing, and customers, and the report must be authored by a qualified independent third party. For USDA B&I, REAP, and Community Facilities submissions above ~$10 million, the demand documentation expected resembles institutional consultancy work product, not a lender summary memo.

The institutional sequence

Eight steps run in order. Three of them concentrate most of the analytical risk.

The first is trade area definition. Four methodologies are relevant for traffic-driven retail. Drive-time isochrone for residential demand. Radial mileage rings for legacy compatibility with appraisal templates. Corridor or AADT capture zone for highway-pad and interstate sites. Mobility-derived true trade areas, generated from anonymized GPS pings by Placer.ai, SafeGraph, or StreetLight Data, calibrated to capture a stated share of observed visit volume. The institutional standard in 2026 layers a true trade area on top of a traditional radial overlay. The two together define the geographic frame.

The second is demand quantification. Three families of methodology. Per-capita or per-household consumption multiplied by trade-area population. ITE trip-generation rates applied to the relevant independent variable. AADT capture method, which dominates traffic-driven retail. Best practice triangulates all three and reconciles to a defensible point estimate with a stated confidence interval, rather than reporting one method as the answer.

The third is capture rate derivation. Capture rate is not a free parameter. It is the share of residual demand the subject realistically commands, derived from a fair-share calculation, calibrated against comparable operating sites, and adjusted for site-specific advantages and disadvantages. A capture rate that exceeds implied fair share without strong site-level justification is the single most common SBA and USDA reviewer flag.

The supporting steps sit between. Demand-driver identification. Demand segmentation across resident, transient, and employment-based pools. Competitive supply netting. Stress testing. Forward-demand projection for structural shifts that change the demand pool itself. The sequence is not optional.

Framework illustration: Demand quantification triangulation across three institutional methods. The interactive emblem above runs three independent demand-quantification methods on a representative site for each asset class: the AADT capture method, the per-capita and per-household consumption method, and the mobility-derived method using Placer.ai true-trade-area data. For a hypothetical suburban gas station at 14,000 one-way AADT, the three methods produce 1.17 million, 3.45 million, and 1.38 million annual gallons respectively, reconciled to a triangulated mid-point of 1.28 million gallons with a stress-tested downside of 950,000. For a 25,000-AADT express car wash with 60,000 residents in a three-mile radius, the methods yield 82,125, 46,800, and 75,000 annual washes, triangulated to roughly 75,000 washes per year at year-three stabilization. For a greenfield truck stop on a high-tonnage I-40 segment with 6,500 AADTT and a 45-mile gap to the next major travel center, the AADTT capture method, the corridor diesel density method drawing on FHWA MF-33SF and HM-20 data, and the ATRI GPS-based mobility method produce 12.45 million, 8.2 million, and 7.5 million annual diesel gallons respectively, reconciling to 9.4 million gallons with explicit autonomous-trucking haircuts in years five through ten. The point of the framework is that no institutional feasibility study should rest on one demand method. Where the three methods converge, the conclusion is defensible. Where they diverge, the divergence itself becomes part of the analytical narrative the lender expects to see documented.

Gas stations and c-stores: the asset class shaped by AADT and EVs

Gas station demand is anchored by AADT. State DOT counts publish two-way Annual Average Daily Traffic; for one-side-of-road sites, which is the typical configuration, the count must be halved and a directional skew adjustment applied. Commute corridors are rarely fifty-fifty by direction. The going-home side captures twenty-five to fifty percent more than the going-to-work side. A study that uses gross two-way AADT without these adjustments doubles its implicit demand pool from the first equation.

Capture rate ranges are well-established in the institutional benchmark literature. Half a percent to two and a half percent of one-way AADT for a stabilized site. The low end describes high-AADT corridors above fifty thousand vehicles per day with heavy competition, off-side-of-road, no signalization, and weak co-tenancy. The median describes suburban arterial pads at fifteen to thirty-five thousand AADT, signalized, branded. The high end describes hard-corner signalized sites on the going-home side with QSR co-tenancy in trade areas with documented under-supply. Outliers above three percent exist in rural monopoly markets with limited substitutes, but they require explicit justification. Site-characteristic adjustment factors are quantitative. Going-home side, plus twenty-five to fifty percent. Signalization, plus fifteen to thirty-five percent. Median barrier preventing left turn, minus twenty-five to fifty percent. These are multipliers on the base capture rate, not free additions.

Productivity benchmarks come from NACS, OPIS, and PDI. The NACS State of the Industry Report pegs the U.S. fueling-location count at roughly 122,600 c-stores selling fuel out of 152,000 total. Industry-average gallons per fueling site sit around three thousand gallons per day, or about 1.1 million gallons per year, but the average is heavily distorted by hypermarkets like Costco and Walmart Murphy that pump four to six times the typical volume. Median independent c-store fuel volume is closer to 1.4 million gallons per year. Foodservice carries 28.5 percent of in-store sales but 38.9 percent of in-store gross profit, which is why inside-sales diversification, not fuel margin, drives c-store profitability.

The EV transition is the structural overlay every post-2024 feasibility study has to model. Consensus forecasts from BLS, S&P Global Mobility, and IEA STEPS converge on EVs reaching forty to fifty percent of new U.S. passenger sales by 2030. Nature and NPJ Sustainable Mobility scenarios place fifty percent of new sales by 2030 and eighty percent by 2035 in the central case. But the in-service vehicle fleet turns over slowly. Average vehicle age sits near twelve years. Even an aggressive EV scenario leaves roughly eighty percent of the fleet ICE-fueled in 2035. The institutional method runs three scenarios. Base case, accelerated EV, delayed EV. Gallons per site decline one to three percent annually in the base case for metro markets, faster in California and Northeast ZEV states, slower in the Sunbelt and rural corridors. The hedge is in-store diversification: foodservice expansion, express car wash co-tenancy, and EV charging infrastructure that monetizes dwell time.

Express car washes: the asset class redefined by subscriptions

The canonical demand equation has not changed: households in the trade area times vehicles per household times annual washes per vehicle times capture rate, cross-checked against AADT times capture rate. What has changed is the revenue model that sits on top of the demand pool. Subscription wash clubs now generate sixty to seventy-five percent of revenue at mature express tunnels, and the methodology has to model the member curve, not just the cars-per-day curve.

Industry-wide structural data anchors the analysis. The U.S. car wash market sits near $20 billion in 2025 with a 5.5 percent CAGR projected through 2035. Roughly 55,000 to 60,000 retail wash establishments operate, with about 200 firms running ten or more locations totaling six thousand sites. Format mix per the ICA 2020 census: in-bay automatic 46.3 percent, tunnel and conveyor 27.9 percent, self-serve 25.8 percent. The format mix is shifting toward express tunnels because the unit economics are better, but the demand methodology has to account for substitution between formats inside any given trade area.

Express tunnel viability thresholds are the gating decision. ICA workshops, Wiggy Wash data, and DRB market saturation analyses converge on a few hard numbers. Minimum population in a three-mile radius runs 25,000 to 35,000 for a stabilized site, with some monopoly small-town markets working below 20,000. Stabilized cars per month above 80,000 typically requires 60,000 to 100,000 vehicle-owning households. Minimum AADT on the fronting street runs 22,000 to 25,000, with 40,000-plus preferred for premium pads. Speed limit must be 45 mph or less, ideally 25 to 45. Lot size 35,000 to 50,000 square feet. The saturation red flag: more than three express tunnels per 50,000 population in a three-mile radius signals oversupply, regardless of how strong the demographics look.

Capture rate is paradoxical in this asset class. Capture decreases as AADT increases, because circulation and stacking become the binding constraint. At 10,000 vehicles per day, express exterior capture runs near 1.7 percent. At 30,000 vehicles per day, 0.8 percent. At 80,000 vehicles per day, 0.5 percent. The institutional underwriting middle is 0.7 to 1.0 percent for a ground-up express tunnel without competition. A peer-reviewed industry analysis demonstrated that capture-rate-from-traffic regressions explain only six percent of variance in actual wash volumes, which is why triangulation against population-based capture, comparable site benchmarks, and Placer.ai true-trade-area validation is institutionally required, not optional.

The subscription dynamic now dominates the model. Mister Car Wash, the publicly traded benchmark, generates roughly seventy-five percent of wash revenue from unlimited wash clubs, with about 2.1 million active members across 500-plus locations. Industry-wide Rinsed quarterly data shows member revenue rising 15 percent year-over-year while retail revenue declines, sometimes sharply when weather amplifies the trend. A mature site benchmark sits near 3,000 active members at $30 monthly ARPU. Member churn runs roughly seven to eight percent. Conversion of retail customers to monthly members runs three percent at immature sites and 13 to 14 percent at mature sites. New site ramps reached 1,000 active members near day 300 in 2023, faster than 2021 or 2022 ramps, because aggressive first-month-free promotions are converting more efficiently than they used to.

A feasibility model that ignores the member ramp curve will systematically over-forecast year one and under-forecast year three. Best practice models months one through thirty-six explicitly, with separate ARPU and churn assumptions for the member pool, and reports the retail-versus-member revenue split at each stage of stabilization. A peak-day capacity check, comparing cars-per-hour throughput against tunnel length, is the supply-side governor that has to be back-tested against the demand forecast. A 120-foot tunnel runs roughly 100 cars per hour stabilized, scaling toward 180 at 140 feet with mature equipment.

Truck stops and travel centers: the asset class defined by corridors, not radii

Truck stop demand is fundamentally different from gas station and car wash demand because it is corridor-oriented, not radius-oriented. The trade area is a band along the interstate, defined by refueling intervals, Hours-of-Service regulation, and the fleet-card economics that route drivers to specific brands. Long-haul trucks with dual 100-to-135-gallon tanks at six and a half to seven miles per gallon have a fuel-tank range of 800 to 1,400 miles, but professional drivers refuel strategically at 500 to 700 mile intervals because that is where the fleet-card discount network places them. The Hours-of-Service rule, codified at 49 CFR Part 395, imposes an eleven-hour daily driving limit within a fourteen-hour shift, a mandatory thirty-minute break after eight cumulative hours, and ten hours off-duty before the next shift. Average daily mileage settles near 500. Mandatory rest creates predictable nightly parking demand peaks between four p.m. and five a.m.

The data infrastructure is federal. FHWA HPMS publishes AADTT, the truck-specific count, separately from the all-vehicle AADT. The Freight Analysis Framework, in its 5.7.1 release through the BTS-ORNL partnership, provides corridor-level freight tonnage and ton-miles with forecasts to 2050. ATRI publishes the annual Top 100 Truck Bottlenecks report and the Operational Costs of Trucking, which benchmarks per-mile cost structure. The Jason's Law Truck Parking Survey, mandated by MAP-21 §1401, documents the systemic parking shortage. ATRI reports one truck parking space per eleven drivers nationally, with 98 percent of surveyed drivers reporting difficulty finding safe parking and 75 percent experiencing the problem at least weekly. Worst hours: four p.m. to five a.m. Worst months: October through February. Worst states: Georgia, Illinois, New Jersey, New York, Pennsylvania. Seventy-nine percent of surveyed truck-stop owners said they have no plans to add parking, which is a structural supply constraint that creates the feasibility opportunity for new entrants.

The demand quantification follows a sequence. Pull state-level diesel gallons from FHWA MF-33SF and MF-21. Pull state interstate mileage from FHWA HM-20. Compute gallons per interstate mile as a flow density proxy. Overlay Jason's Law parking shortage data, ATRI Top 100 bottleneck adjacency, and FAF5 corridor tonnage. States with high gallons per mile, parking deficit, and bottleneck cluster identify the highest unmet-demand prioritization, which currently maps onto New Jersey, Texas, Georgia, Florida, California, and Illinois. Distance to the next major travel center governs site-level capture. Institutional rule-of-thumb is one major travel center every twenty-five to sixty miles on heavily-trafficked Class 1 freight corridors, sparser on secondary interstates.

Travel center economics are diversified by design. NATSO describes the modern travel center as five to six stores in one. Diesel fuel desk and diesel lanes. Convenience store. Multiple QSR co-tenants, typically two to four. Often a full-service restaurant under a heritage brand. Showers, ten to thirty per major site. Truck parking, fifty to five hundred spaces. Truck and tire repair facilities. CAT scales. Laundry, lounges, gaming. DEF, Diesel Exhaust Fluid, at every modern diesel pump because every post-2010 EPA-compliant heavy-duty diesel engine requires it. Margin profile: fuel is thin at three to seven cents per gallon retail, inside merchandise runs 25 to 55 percent, QSR generates rent and royalty income, and the repair shop sits at 30 to 50 percent labor margin. Diesel concentration on the fuel side runs 85 percent typical. TravelCenters of America reported eighty percent of total revenue from fuel before its acquisition by BP. The chain hierarchy is well-defined: Pilot Flying J at roughly 750 locations, Love's Travel Stops at roughly 600, TA-Petro at 280, with an ecosystem of 800-plus independents competing on price and corridor niche.

The autonomous trucking overlay is now methodologically required for any project with a ten-year horizon. Aurora completed driverless 1,200-mile Dallas-to-Houston runs in 2024. Aurora, Kodiak, Torc Robotics, MAN, and Scania all plan Level 4 commercial deployments by 2030. Industry forecasts converge on roughly 500,000 autonomous trucks on U.S. roads by 2030, with up to fifty percent of freight movement automated by 2040. The methodological implication is asymmetric. Autonomous trucks reduce HOS-driven parking demand because they do not require ten-hour driver rest periods. They reduce QSR and full-service restaurant amenity demand because there is no driver to feed. They may marginally increase fuel-only stop frequency. Diesel volume per stop is largely unchanged. The institutional method haircuts amenity revenue by ten to twenty percent in years five through ten in the base case, and by thirty to forty percent in the accelerated-autonomy scenario.

What gets stress tested

Three variables drive every traffic-driven retail pro forma: capture rate, average ticket or per-unit margin, and ramp timing to stabilization. The institutional stress matrix runs all three in parallel.

Base case at stated capture, stated ticket, stated ramp, with DSCR at or above 1.25 times. Downside at minus twenty-five percent capture, minus ten percent ticket, ramp slowed by six months, with DSCR at or above 1.10 times. Severe downside at minus forty percent capture, minus fifteen percent ticket, ramp slowed by twelve months, with DSCR at or above 1.00 times. Asset-specific stresses layer on top. For gas stations, accelerated EV adoption removing twenty percent of gallons by 2030 instead of base case ten percent. For car washes, member churn rising to twelve percent and conversion falling to four percent. For truck stops, autonomous deployment accelerated five years and amenity revenue cut thirty-five percent.

,

Year-one ramp curves are asset-specific. Gas stations reach seventy to eighty percent of stabilized in year one, ninety to ninety-five in year two, and one hundred percent in year three for branded sites at proven trade areas. Express car washes reach fifty to sixty-five percent in year one because the member ramp dominates, seventy-five to eighty-five in year two, and ninety-five-plus in year three. Truck stops reach sixty to seventy-five percent in year one because trucker network awareness and fleet-card relationships take six to twelve months to build, eighty to ninety in year two, and ninety-five-plus in year three.

Tornado-style sensitivity reports rank the variables that most affect DSCR. The institutional study identifies which two variables dominate and demonstrates explicit acknowledgement of the dominant exposure to the lender, CDC, and SBA or USDA reviewer.

Interactive tool: Capture rate calibration by asset class with site-characteristic adjustment factors. The emblem above lets the reader set the AADT for the subject site and toggle the site-specific factors that drive capture rate, returning a calibrated range in real time. Base capture rate ranges follow institutional benchmarks. Gas stations run from 0.4 to 2.5 percent of one-way AADT, declining as traffic volumes rise. Express car washes invert the relationship because circulation and stacking become the binding constraint at higher traffic counts: 1.2 to 1.7 percent of AADT below 15,000 vehicles per day, falling to 0.4 to 0.7 percent above 60,000. Truck stops vary from 1.5 to 5.0 percent of AADTT depending on AADTT band and competitive density. Site-characteristic adjustment factors apply quantitative multipliers to the base range. For gas stations, the going-home side commands a 25 percent capture premium, signalization adds 20 percent, hard-corner two-frontage sites add 15 percent, and a median barrier preventing left turns subtracts 35 percent. For express car washes, population above 35,000 in a three-mile radius adds 20 percent, no competing tunnel within the trade area adds 25 percent, and saturation above three tunnels per 50,000 residents subtracts 30 percent. For truck stops, distance of 40 miles or more to the next major travel center adds 30 percent, ATRI Top 100 bottleneck adjacency adds 20 percent, Jason's Law parking deficit adds 18 percent, and a branded competitor within 15 miles subtracts 25 percent. The cumulative multiplier is the product of every active factor, applied to the base range. The output is the calibrated capture rate range a lender, CDC, or USDA reviewer can audit step by step. The framework makes a single point: capture rate is derived, not assumed, and a study that cannot show the multipliers cannot defend the number.

Where the methodology is heading

Five directional shifts are reshaping institutional demand analysis for traffic-driven retail.

Mobility data is now the validation standard, not a supplementary input. Placer.ai true trade area, StreetLight FHWA-validated AADT outperforming forty-eight-hour temporary counts on mean absolute percentage error, SafeGraph and Dewey POI-level visit panels. SBA and USDA reviewers in 2026 increasingly expect this layer in premium submissions, particularly for new-build feasibility above $5 million.

EV-adjusted gas station forecasting is now expected for any post-2024 feasibility submission. Multi-scenario gallons-per-site projections drawing on Nature and NPJ Sustainable Mobility scenarios, IEA STEPS, and S&P Global Mobility forecasts have replaced single-point trajectories. The institutional default models base case, accelerated EV, and delayed EV in parallel.

Subscription-driven demand modeling has fundamentally restructured express car wash methodology. Member ARPU plus churn rate plus conversion rate stochastic models now replace simple cars-per-day pro formas. Rinsed quarterly data, covering roughly 3,500 locations, has emerged as the de facto industry benchmark for member economics.

Autonomous-trucking demand impact is now methodologically required for truck-stop feasibility on ten-year-plus horizons. The 2027 to 2030 commercial deployment window forces explicit scenario modeling.

The regulatory regime has tightened. SBA SOP 50 10 8 eliminated lender discretion in 2025. USDA OneRD harmonized rural lending under 7 CFR Part 5001. NAICS 484 was added to the environmentally sensitive list, pulling truck stops into Phase I ESA scope. Each change raises the standard on the demand chapter specifically, because that is where reviewers focus first.

Closing observation

There is no asset class where demand analysis is optional, and no asset class where it can be performed without ground-truthing the trade area, the comparable sites, and the corridor. The studies that secure loan approval, and the studies that survive the operating reality two and three years later, share the same structure. A defensible trade area defined by the right methodology for the asset class. A demand pool quantified by triangulation across at least two independent methods. Competitive supply explicitly netted out before capture is calculated. A capture rate derived from fair share and adjusted for site-specific factors that are documented and quantitative. A forward-demand projection that acknowledges structural shifts the asset class is undergoing. And a sensitivity analysis that tells the lender precisely what the project must clear, not merely what the base case projects. Under the post-2025 regime, anything less is no longer institutionally acceptable.

April 29, 2026 by Michal Mohelsky, J.D.

MMCG Invest, LLC is feasibility study consultant that provides independent, third-party feasibility studies for SBA and USDA guaranteed loan programs across all commercial real estate asset classes, including multifamily, hotel, industrial, retail, self-storage, senior living, and mixed-use properties. Our studies incorporate absorption rate analysis, lease-up modeling, pre-stabilization cash flow bridging, pro forma and scenario-based stress testing to meet the analytical rigor required by leading government-guaranteed lenders, CDCs, and institutional investors. For more information, contact our team directly.

Evaluating a development or acquisition that requires defensible absorption assumptions? Reach out to discuss how our methodology supports your lending decision.

Michal Mohelsky, J.D. | Principal | mmcginvest.com 

Contact: michal@mmcginvest.com

Phone: (628) 225-1125

Disclaimer: This report is provided for informational purposes only and does not constitute investment advice. Data presented herein is derived from proprietary MMCG databases and third-party sources believed to be reliable; however, MMCG Invest makes no representation as to the accuracy or completeness of such information. Figures from third-party industry databases have been independently verified and, where appropriate, adjusted to reflect MMCG's proprietary analytical methodology. Past performance is not indicative of future results.

Sources

1. U.S. Small Business Administration. Standard Operating Procedure 50 10 8: Lender and Development Company Loan Programs, effective June 1, 2025; Appendix 7 (Gas Station Loans). https://www.sba.gov/document/sop-50-10-lender-development-company-loan-programs

2. U.S. Department of Agriculture, Rural Development. 7 CFR Part 5001: OneRD Guarantee Loan Initiative(B&I, REAP, Community Facilities, Water and Waste Disposal), Appendix A to Subpart D. https://www.rd.usda.gov/onerdguarantee

3. Appraisal Institute. The Appraisal of Real Estate, 15th edition, and Market Analysis for Real Estate (Stephen F. Fanning, MAI). The Six-Step Process anchoring demand and capture analysis.

4. Federal Highway Administration. Highway Performance Monitoring System (HPMS); Freight Analysis Framework FAF5.7.1 (BTS-ORNL); Jason's Law Truck Parking Survey (2018-2019, 2026 update); Truck Parking Development Handbook (2022); MF-33SF, MF-21, and HM-20 fuel and mileage data. https://www.fhwa.dot.gov

5. Institute of Transportation Engineers. Trip Generation Manual, 11th edition. Land Use Codes 944, 945, 947, 948, 950, 960 covering gas/c-store, car wash, and truck stop.

6. NACS (National Association of Convenience Stores). State of the Industry Report (2025 data released early 2026); NACS-NIQ TDLinx store count; Convenience Stores Fact Sheets. https://www.convenience.org

7. International Carwash Association. Industry census, consumer surveys, and operating standards. Supplemented by Rinsed Quarterly Industry Report for subscription-driven member economics. https://www.carwash.org

8. NATSO and ATRI (American Transportation Research Institute). Travel Center Trends and Insights (NATSO Foundation, annual); ATRI Top 100 Truck Bottlenecks; Operational Costs of Trucking; Critical Truck Parking studies. https://www.natso.com · https://truckingresearch.org