S09.7 · Retail & Commerce
An $817.5B US category pivoting to foodservice as fuel margins erode, with limited direct AI exposure versus EV adoption risk.
US convenience retail took in $817.5B across in-store and fuel in 2025 — in-store sales of $341.2B, up 1.7% year over year (NACS, 2026). Fuel sales fell roughly 5.4% on lower gas prices, leaving in-store and foodservice as the segment's real growth engine while fuel margins structurally erode with EV adoption. Of all the segments in this sector, convenience has the least direct exposure to agentic AI. The long-term threat that matters is EVs thinning forecourt traffic, not AI-driven comparison shopping.
The data is dominated by the US market (NACS), though localized real-estate-driven businesses exist globally — UK forecourts and their European equivalents among them. Large chains such as 7-Eleven, Circle K/Couche-Tard and Casey's are consolidating a still-fragmented, heavily franchised and independent base: 151,975 US stores in 2025, down 0.2%, even as fuel-selling locations hit an eight-year high. Fuel wholesalers, refiners and food distributors sit upstream; convenience-driven consumers making impulse and fill-in purchases sit downstream. Fuel margins are thin, in-store margins healthier, and the business is site-selection-driven, with regulatory gating around fuel storage and environmental compliance.
Store count edging down while fuel-selling locations hit an eight-year high says where the segment's real competition lives: this is a real-estate and traffic-capture business first and a merchandising business second. Site quality sets fuel volume, fuel volume sets footfall, and footfall is what the higher-margin in-store and foodservice lines are built to monetize. That sequencing is why operators are pulling the foodservice lever, not the fuel-pricing lever, as fuel margins compress structurally with EV adoption.
Foodservice build-out replaces fuel margin that is going away; investment in EV charging infrastructure is optionality on future forecourt traffic as the vehicle mix shifts — spend to stay in the traffic business under a different power source. Direct AI exposure stays limited because impulse purchases are not spec-comparable the way a commodity electronics purchase is; a shopping agent has little to optimize when the transaction is a coffee or a candy bar bought on the way past the register.
AI earns its keep in the back office instead. Dynamic fuel pricing and labor-scheduling automation are compressing site-level operating expense — a straightforward efficiency gain, not a competitive threat. The foodservice pivot is running now; the larger structural risk, EV adoption reducing the number of vehicles that need to stop for fuel at all, plays out on a longer, 5-10 year horizon.