Drive-Thru Speed of Service: The Operator's Guide
Kyler Close
Drive-thru speed of service is the time a guest spends moving through your lane, from arriving at the order point to leaving the pickup window with a complete order. It matters because every extra second hits throughput, guest satisfaction, and the unit economics of your busiest channel.
What Is Drive-Thru Speed of Service, and Why Does It Matter?
Drive-thru speed of service (SOS) is the elapsed time between a guest reaching your lane and driving away with the right order, and it is the single most visible measure of how well a unit runs.
Operators track it because the drive-thru is where operational problems show up first. A short-staffed shift, a slow grill, or a confusing lane layout all become seconds on the clock. Those seconds turn into abandoned orders, fewer cars per hour, and lower AUV.
Guests also notice. Savi's Drive-Thru Disruptors research, which analyzed more than 250,000 customer reviews, found that drive-thru sentiment influences 73% of a restaurant's overall review score. Speed is not only an operations metric. It shapes how the whole brand is rated.
Four things make SOS hard to manage at scale:
Averages hide the worst moments, such as a lunch rush that runs long while the daily number looks fine.
Each unit has its own layout, staffing, and equipment, so the same standard plays out differently everywhere.
Managers often learn about slow service from complaints, not from data.
Without a consistent way to time every car, two locations can report the same number and mean different things.
How Do You Measure Drive-Thru Speed of Service?
You measure it by timing each car at defined points in the lane and rolling those timestamps into a total time and a set of segment times.
Most operators break the lane into the same few segments. Looking at each one separately tells you where the time actually goes:
Greet time: how long a guest waits at the order point before a team member responds.
Order time: how long the guest takes to place the order and confirm it.
Service time: how long the car waits between the order point and the pickup window.
Window time: how long it takes to pay and receive the order.
Total time: the full path from arrival to departure.
The tool you use to capture those timestamps shapes what you can learn. Traditional systems rely on in-ground sensors, and a clear explanation of how loop timers compare with camera timers helps you decide which fits your sites. Loops can be accurate, but they require installation at every unit, and they tell you when a car crossed a point, not why it stalled.
Camera-based timing reads the same lane using video you may already be recording. Savi's Drive-Thru Flow analytics for lane timing time the lane from video so managers can see where service slows.
The hardware difference matters in dollars. Swig avoided an estimated $4,000 to $6,000 per location in loop-timer installs, an estimated $1.1M for its franchisees. In the words of its COO, Swig got "ground-breaking insights without breaking ground at any of our sites."
How Do AI and Computer Vision Time the Drive-Thru?
AI and computer vision time the drive-thru by detecting vehicles and their position in the lane from video, so timestamps come from what the camera sees rather than from a sensor.
That is a true computer vision task: the system recognizes a car arriving, tracks it through the lane, and logs when it reaches each point. No transaction is needed to trigger it. A walkthrough of how computer vision improves drive-thru times shows how that detection turns raw footage into lane-level measurements.
Timing is only the starting point. A closer look at AI-driven drive-thru timing analytics explains how those measurements become patterns an operator can act on, such as which positions in the lane back up first and which dayparts slow down.
There is also a lot of noise in this category. Plenty of vendors pitch drive-thru AI as a voice-ordering robot or a futuristic overhaul. Savi's take on practical drive-thru AI without the sci-fi argues for the opposite: start with the video you already have and use it to answer ordinary operating questions.
When evaluating any drive-thru AI claim, ask three things:
Does it work with the cameras already installed, or does it require new hardware at every site?
Does it report segment times, or only a single total?
Can a general manager read the output without a data analyst?
What Are Stack, Pull-Forward, and Drive-Off?
Stack, pull-forward, and drive-off are the three lane events that tell you a drive-thru is losing money even when the average time looks acceptable.
Each one describes a different kind of breakdown, and each is worth tracking on its own:
Stack: cars queue up behind the order point or window faster than the team can clear them, which signals a throughput problem at the bottleneck.
Pull-forward: a team member asks a guest to park and wait for a late order so the lane can keep moving. It protects the timer but hurts the guest experience.
Drive-off: a guest leaves the lane before ordering or before reaching the window, which is lost revenue that never shows up in a speed report.
The danger is that a pull-forward can make a time look better than the service really was. If you only report average total time, a unit can hit its target while guests wait in parking spaces. Counting stacks, pull-forwards, and drive-offs alongside SOS gives you a more honest picture of lane health.
How Do You Benchmark Speed of Service Across Locations?
You benchmark by comparing every unit on the same definition of time, at the same lane points, within the same daypart.
Without that consistency, ranking units is guesswork. A detailed breakdown of speed of service benchmarks for QSR covers how to set targets that account for volume, layout, and menu.
The harder part is doing this at scale. Managing drive-thru speed across multiple locations means every district manager sees the same numbers, the same way, without waiting for a monthly report or a spreadsheet from each unit.
A useful multi-unit benchmark views results in three ways:
Unit against its own history, to show whether it is improving.
Unit against peers with similar volume and layout, to show who has room to move.
Unit against the brand standard, to show where coaching is needed.
Savi's results with Swig show what this looks like in practice. Swig improved its drive-thru speed 7–10% within the first 90 days, with its COO noting, "Last month we had our fastest drive-thru speeds ever."
Why Does Historical Data Matter More Than Real-Time Alerts?
Historical data matters more because the causes of slow service repeat on a schedule, and only a long record lets you see the pattern.
Real-time alerts are useful for a manager on shift. They tell you a lane is backing up right now. But they do not explain why the same unit backs up every Friday at 12:15.
A case for using historical data to dominate the drive-thru argues that trend data is where coaching decisions come from. With weeks of lane history, you can tell the difference between a one-off bad shift and a staffing gap that shows up every week.
Historical data answers questions real-time views cannot:
Which dayparts consistently miss target, and by how much?
Did a menu change, a new hire class, or an equipment swap change lane times?
Which units improved after coaching, and which did not?
How Do Speed of Service and Labor Vary by Daypart?
Speed of service and labor are tightly linked, and the link changes by daypart, because the same crew size handles very different volumes at breakfast, lunch, and dinner.
A unit that looks fine over a full day can still be understaffed at lunch and overstaffed at 2 p.m. Looking at SOS by daypart exposes those gaps. It lets you move hours to where they matter instead of adding labor across the board.
When reviewing labor against speed, operators check:
Whether the lunch and dinner peaks have enough team members on the order point and window positions.
Whether the slowest segment lines up with a specific station or role.
Whether new team members are scheduled during the heaviest volume without support.
Whether slow stretches follow shift changes or breaks.
The goal is not to cut hours. It is to put the right people in the right position at the right time, then confirm in the data that service improved.
How Do You Protect the ROI of a Drive-Thru Endcap or New Build?
You protect the ROI of a new drive-thru by measuring lane performance from day one, so you know within weeks whether the investment is paying back.
Endcaps and new drive-thru builds are expensive, and the sales case assumes the lane will perform. If speed lags, the projected throughput never materializes. A look at de-risking a drive-thru endcap with cloud analytics shows how early data helps you catch layout or staffing issues before they erode returns.
For a franchisee evaluating a new site, early visibility answers practical questions. Is the lane layout causing stacks? Is the order point or the window the constraint? Is the unit hitting its targets by the end of the first quarter?
What Else Can Computer Vision Do in a Restaurant?
Computer vision can do much more than time the lane: it detects behavior and visual patterns from video that point to guest, compliance, and loss issues.
An overview of computer vision for QSR restaurants maps the full set of uses. A companion explanation of AI cameras for quick-service restaurants covers what the camera hardware needs to do and what can run on the cameras you already own.
The value falls into three areas, in this order:
Guest engagement and service times: detecting how long guests wait, how lanes and counters flow, and where service slows down.
Brand compliance: spotting visual deviations from the brand standard, such as uniform, cleanliness, and station setup patterns.
Loss prevention: surfacing unusual behavior and baseline deviations that a manager would never catch by scrubbing footage.
The drive-thru is the clearest entry point because the lane gives you a clean, repeatable thing to measure. Once that works, the same video supports the other two areas.
How Do You Choose a Video Analytics Platform for Your Drive-Thru?
Choose a platform that works with your existing cameras, reports across every location, and supports more than drive-thru timing from the same video.
Start with a clear view of what QSR video analytics solutions actually do, since vendors use the term for very different products. Then compare options using a structured guide to the best AI video analytics platforms for QSR chains.
If you already have a system and are unhappy with it, a review of alternatives to legacy QSR video analytics lays out what to look for in a replacement. The most overlooked part of any decision is what you already own. Your cameras are an asset, and the case for using the cameras already in your locations explains why rip-and-replace is not required.
Here is the architectural point that should drive the decision. The same cloud video can support IT and loss prevention, not just drive-thru timing. A cloud-based foundation makes it easier to add new uses over time. As computer vision and AI advance, a cloud-architected foundation lets you adopt new tools without ripping out on-site infrastructure. Treat it as a foundation decision, not a point-solution purchase.
Key Takeaways
Drive-thru speed of service is best managed by segment (greet, order, service, window), not by a single total time.
Stack, pull-forward, and drive-off keep average times honest and show lost revenue that speed reports miss.
Consistent definitions across units are what make benchmarking and multi-location management work.
Historical data and daypart views reveal repeating causes that real-time alerts cannot explain.
Your existing cameras can time the lane, and the same cloud video can support IT and loss prevention.
See what your own lanes are telling you, one location at a time or across the whole brand. Request a Savi demo
Frequently Asked Questions
What is a good drive-thru speed of service?
A good target depends on your menu, volume, and daypart, so there is no single number that fits every brand. Set targets by comparing each unit to its own history and to peers with similar layouts. Track segment times, not just the total, so you know where to improve.
How do you improve drive-thru speed of service?
Start by measuring each lane segment to find the real bottleneck. Then coach on that segment, adjust labor by daypart, and check the data weekly to confirm the change worked. Small fixes at the order point or window often move the total more than a large operational overhaul.
Do I need new cameras or sensors to time my drive-thru?
No. Camera-based timing can use the cameras you already have, which avoids installing a loop sensor at every location. A site visit or camera review can confirm whether your current placement covers the lane.
What is the difference between a loop timer and a camera timer?
A loop timer uses an in-ground sensor to log when a car crosses a point in the lane. A camera timer uses video to detect the car and its position, so it can report more segments and also show what happened in the lane. The tradeoff is hardware installation at each site versus using video you already record.
Can drive-thru analytics work across hundreds of locations?
Yes, if the platform is cloud-based and uses one consistent definition of time at every unit. That lets district managers and executives compare sites, daypart by daypart, without manual reports. Consistency is the key requirement, because inconsistent data makes unit rankings meaningless.
