Drive-Thru Site Selection: The Criteria That Decide Throughput

For a modern QSR or fast-casual brand, the drive-thru is not an amenity attached to a restaurant. It is the restaurant. When 65 to 85 percent of sales come through the window, throughput is the product, and the site either delivers it or quietly caps your revenue for the length of the lease. That is the part standard site criteria tend to underweight: a trade area can be perfect and a parcel can still be incapable of hitting plan because it cannot physically hold the cars.

Most drive-thru misses aren’t demand misses. They are capacity and access misses. The demand showed up, got to the entrance, saw eleven cars, and kept driving. Below is the set of criteria that decides whether a drive-thru site can convert the traffic it’s underwritten against, and how to get those constraints into the forecast rather than discovering them in year two.

Stacking Depth Is a Revenue Cap, Not a Design Preference

Everything else on this list is important. Stacking is arithmetic. Your peak-hour sales cannot exceed service rate multiplied by the hour, and your realized sales cannot exceed the share of arriving customers willing to join the queue they can see from the street. Once the visible queue passes a customer’s tolerance, they don’t wait. They leave, and the sale goes to whoever is a quarter mile down the road.

So the requirement should be derived, not inherited from a prototype sheet. Take the peak-hour transaction count from your strongest comparable stores, divide by an honest service time, and translate the result into vehicles that must fit on the parcel between the entry and the pickup window. Then check that queue against the actual site plan.

• Depth behind the order point is what protects the peak; depth between order point and window only smooths service.
• Spillover matters legally, not just operationally. Many jurisdictions require the queue to be contained on site, and a queue that reaches the right-of-way can trigger enforcement or an unhappy landlord.
• The parking field is not overflow. A queue that wraps through parking blocks the dining-room customers and any mobile pickup traffic you were counting on.
• Peak, not average. A lot that handles the daily mean and fails the 7:30–9:00 window fails at exactly the hour that pays for the deal.

Ingress, Egress, and the Right-Turn-In Advantage

A drive-thru captures traffic that is already in motion. Every additional decision, lane change, or gap in oncoming traffic that a customer has to negotiate reduces capture. That is why the right-turn-in is worth real money: the customer decelerates, turns, and is on your lot with no conflict. A left-turn-in across two or three lanes at 45 mph is a different transaction entirely, and the friction shows up as a lower capture rate rather than as anything visible on a Tuesday-afternoon tour.

Diligence should cover the full access picture: curb-cut location relative to the functional area of the intersection, median breaks and whether they are permanent, DOT access-management standards, deceleration and turn lanes, signal timing, and any shared-access or cross-easement language that lets a neighbor constrain your circulation. Egress deserves equal attention. An exit that forces a left turn into a signal queue at 5:15 p.m. makes the visit feel expensive, and repeat frequency, which is the whole economics of a coffee or breakfast concept, is what erodes.

Side of Street, Commute Direction, and Daypart

This is where the same parcel changes value depending on which concept is looking at it. A coffee or breakfast drive-thru is fundamentally an inbound-commute business: it wants to be on the right-hand side of the road heading toward employment concentration, in the morning, with an easy in and easy out. A dinner-weighted concept wants the outbound side, on the way home, where the trip is less time-pressured and the basket is larger.

Treating a traffic count as one number destroys this distinction. Twenty thousand cars a day split heavily toward you in the morning is a completely different asset from twenty thousand split against you. Directional splits by daypart, layered against where the residential base sleeps and where the employment base works, are what make the difference legible. That is the same reasoning behind daytime versus nighttime population analysis and why a rigorous trade-area analysis for a drive-thru looks less like a ring and more like a set of directional flows.

Lot Geometry: Escape Lanes, Dual Lanes, Order-Point Placement

Two parcels with identical acreage and identical traffic can differ by a wide margin in throughput because of shape. Depth, frontage-to-depth ratio, corner versus mid-block position, existing easements, detention requirements, and setback lines all determine whether a compliant, high-capacity circulation pattern actually fits.

The escape lane

A bypass lane lets a customer leave the queue after ordering, and lets an order that needs extra time pull forward without stopping the line. On a constrained lot it is the first thing value-engineered out, and it is usually the wrong thing to cut, because a single long ticket in a single-lane configuration halts every car behind it.

Dual order points and tandem lanes

Side-by-side lanes that merge before the window can raise order-taking capacity substantially, but they demand width, a clean merge, and enough post-merge stacking that the merge itself doesn’t become the choke point. A dual-lane prototype forced onto a parcel that can’t support the merge geometry often performs worse than a well-designed single lane.

Where the order point sits

Order-point placement decides how much of your stacking is usable. Push it too close to the entrance and the queue backs into the street; push it too far and you lose the pre-order buffer that keeps the kitchen loaded. It also needs a clear line of sight from the entry so an arriving customer can read the queue length and commit.

Visibility and Decision Distance at Speed

A drive-thru customer needs to recognize the brand, decide, signal, and decelerate. At 25 mph that sequence takes a short stretch of frontage. At 50 mph on a divided arterial it takes far more, and a monument sign hidden behind a pylon, a mature tree line, or a grade change can cost you the decision entirely. Judge visibility from the driver’s seat, at the prevailing speed, in both directions, and in the season when the canopy is full.

Sign rights belong in the same conversation. Signage height limits, monument-sign restrictions in the ordinance, and whatever the landlord actually controls in a multi-tenant pad situation can quietly undercut a site that looked visible on the tour. So can a future pad that hasn’t been built yet on the parcel in front of you.

Zoning, Moratoriums, and the Diligence You Do First

Drive-thru entitlement has become a genuine constraint on growth. Municipalities have adopted bans, unit caps, conditional-use requirements, queue-containment standards, and design overlays, frequently citing congestion, pedestrian safety, and emissions. Some downtowns and corridors are simply closed to new drive-thrus. Others allow them only with a public hearing, which introduces both delay and political risk.

This belongs in the first pass of market screening, not in late-stage diligence. Read the ordinance and the overlay map, scan recent planning-commission agendas and minutes for the submarket, and note where drive-thrus were approved, conditioned, or denied. A market where the last three drive-thru applications died in hearing is a market where your pipeline assumptions need rewriting. Handled early, entitlement risk is a filter; handled late, it is a lost year.

Mobile Order Pickup Is Reshaping the Lot

Digital ordering added a second physical channel to a lot that was designed for one. Mobile pickup wants dedicated stalls near the door, a walk-up window or vestibule, courier staging, and a route to all of it that doesn’t require entering the drive-thru queue. Where that route doesn’t exist, the two channels fight: couriers park in the escape lane, pickup cars back up at the entrance, and both experiences degrade at peak.

The practical implication for site criteria is that acreage and circulation requirements have gone up even as dining rooms have shrunk. Newer prototypes lean into it, with dedicated pickup lanes, order-ahead bypasses, and in some cases no interior seating at all. When you write or refresh your criteria, specify pickup capacity explicitly alongside stacking depth rather than assuming the parking field will absorb it.

Put Throughput in the Model, Not Just on the Tour

Here is the modeling point that ties all of it together. Most revenue forecasts are demand-side: population, daytime employment, traffic, competition, analog performance. Applied to a drive-thru on a constrained lot, that kind of model will over-predict, sometimes badly, because the demand it counts cannot physically be served. The site can hold six cars; the forecast assumed the customers who would have been the seventh through the fourteenth.

The fix is to treat capacity as an explicit term. Estimate peak-hour arrivals, estimate the site’s achievable service rate given stacking, lane configuration, and access, and cap or haircut the forecast where the two collide. Done properly this changes deal decisions: it turns two parcels that scored identically on demographics into a clear yes and a clear no, and it tells you what a second lane or ten more feet of stacking is actually worth in dollars during lease negotiation. It also disciplines the pipeline, because the sites that fail on throughput fail early and cheaply.

This is the discipline behind Locate’s approach to new-store sales forecasting: model the revenue a specific site can realistically deliver, constraints included, and keep the analysis attached to the people negotiating the deal. For the broader criteria set beyond the window, see our guide to restaurant site selection. If you want a second read on a drive-thru parcel before you commit, or a throughput-aware forecast for a market you’re entering, talk to our team.

Common Questions

How many cars should a drive-thru be able to stack?

It depends on your peak and your service time, not on a universal number. The math is simple: cars stacked behind the order point are the only buffer you have between a rush and a drive-off. Coffee and beverage-led concepts with 30-second service windows and a violent morning peak often want a dozen or more vehicles of stacking plus an escape lane; a burger or chicken concept with a longer service time but a flatter curve may function on fewer. The right way to set the requirement is to take your busiest analog store's peak-hour car count, divide by your realistic service rate, and see how deep the queue gets before the lot fills. If the site can't hold that queue without spilling into the parking field or the public right-of-way, your peak sales are capped no matter how good the trade area is.

Why does side of the street matter so much for a drive-thru?

Because a drive-thru sells to traffic that is already going somewhere, and the direction of that traffic changes by daypart. A coffee concept lives on the morning inbound commute, so it needs to be on the right-hand side of the road heading toward employment centers, ideally with a right-turn-in that requires no crossing of traffic. A dinner-weighted concept wants the evening outbound side, on the way home. The same parcel can be excellent for one concept and mediocre for the other. Counting total daily traffic without splitting it by direction and daypart is one of the most common ways a drive-thru forecast goes wrong.

What ingress and egress problems should kill a drive-thru deal?

The ones that force customers to work. A left-turn-in across multiple lanes with no median break, a curb cut inside the functional area of a signalized intersection, a raised median that makes the site accessible from only one direction, an egress that dumps cars into a queue at a red light, or a shared access easement controlled by a neighbor who can restrict circulation. Each of these shaves conversion off passing traffic in ways that are hard to see on a site visit but very visible in sales. Get a traffic engineer and the DOT's access-management rules involved before you get emotionally committed to a parcel.

Are drive-thru moratoriums a real risk in site selection?

Yes, and they are worth diligence in the very first week of looking at a market. A growing number of cities and towns have adopted outright bans, caps, conditional-use requirements, or design standards on new drive-thru facilities, often citing congestion, pedestrian safety, or emissions. Others allow drive-thrus only in specific overlay districts or require a public hearing that can add many months. Screen the zoning ordinance, the overlay map, and recent planning-commission agendas for the submarket before you spend money on a site, because entitlement risk is the fastest way to lose a year of pipeline.

How do mobile orders change drive-thru lot requirements?

They add a second demand stream that competes for the same asphalt. Digital pickup means curbside stalls, a walk-up window or vestibule, staging space for couriers, and a circulation pattern that lets a pickup car reach a stall without entering the drive-thru queue. Concepts that retrofitted mobile order onto a lot designed for a single lane often discover the two channels interfere: couriers block the escape lane, or pickup traffic clogs the entry. New site criteria should specify pickup stalls and a separate path for them, not just stacking depth.