- $915M
QSR AI market size in 2024
- $12B
Projected market by 2034
- 22 s
Seconds cut by Wendy's FreshAI
- 650+
Taco Bell stores with voice AI
The fast-food drive-thru was born in 1947. For nearly 80 years, the core interaction remained essentially the same: a customer talks to a human through a speaker box, places an order, and pulls forward to pay. It was simple, reliable, and deeply resistant to change. But in the span of just 18 months, every major QSR chain has either deployed or announced an AI-powered ordering system. The speaker box is learning to talk back -- and it is only the beginning.
This is not just a story about drive-thrus getting smarter. The real transformation extends far beyond the speaker box -- into WhatsApp conversations, in-app chat, and text messaging. Welcome to the era of the Chat-Thru: where your next burger order might start with a text message instead of rolling down your window.
The AI Drive-Thru Arms Race: Who's Doing What
The numbers tell the story of an industry in full sprint. According to Qu's State of Digital report, 51% of limited-service brands are actively investing in AI, with another 22% planning to start this year. A Popmenu survey found that 74% of restaurant operators expect AI to become a staple technology within three years. This is not experimentation anymore -- it is a competitive necessity.
From McDonald's massive Google Cloud partnership to Wendy's 160-location FreshAI rollout, each chain is taking a different approach -- with wildly different results. Here is how the timeline has unfolded.
The Fast-Food AI Timeline
2023
McDonald's Launches IBM AOT Pilot
Automated Order Taking system deployed across select U.S. drive-thru locations in partnership with IBM.
Jul 2024
McDonald's Ends IBM Partnership
Pilot terminated after persistent order errors, including a viral 260-McNuggets incident that went viral on social media.
Early 2025
Wendy's FreshAI Hits 160+ Locations
Google Cloud-powered system reports 99% accuracy and 22-second wait time reduction. Spanish language support added.
Mar 2025
Taco Bell Deploys in 650+ Stores
Omilia Voice AI with Nvidia partnership rolled out across U.S. locations. Later scaled back to a 'sometimes tool' approach.
2025
McDonald's Pivots to Google Cloud
Major tech overhaul across 43,000 restaurants. AI-powered Accuracy Scales deployed in thousands of drive-thrus in 12 markets.
Jan 2026
Papa John's Launches Google Gemini Agent
First brand to deploy Google Cloud Food Ordering agent. Features include Intelligent Deal Wizard and no-tap reordering.
2026
Burger King's 'Patty' Goes Live
OpenAI-powered chatbot monitors drive-thru conversations across 500 U.S. pilot locations. Nationwide rollout planned by year-end.
Report Card: How AI Is Actually Performing
Headlines love the success stories, but the real picture is more nuanced. According to Intouch Insight's 2025 Emerging Experiences study, which evaluated 120 AI-processed orders, AI drive-thru service is faster than humans but less accurate. The AI averaged 3 minutes and 53 seconds from order start to food in hand, compared to 4 minutes and 15 seconds for human order-takers. Speed is the clear win.
But accuracy tells a different story. AI order accuracy came in at 83% versus 87% for humans. Perhaps most telling: 22% of AI orders still required human intervention, and 65% of AI order issues stemmed from customization failures -- the 'no pickles, extra mayo' requests that trip up even the most advanced natural language models.
AI vs. Human: Drive-Thru Performance
AI Service Time
3:53
Average time from order start to food in hand
Human Service Time
4:15
Average time with human order-taker
AI Order Accuracy
83%
Correct items delivered per Intouch Insight 2025
Human Order Accuracy
87%
Human order-takers still lead on accuracy
The notable outlier is Wendy's FreshAI, which claims 99% order accuracy and a 22-second reduction in average wait times across its 160+ deployed locations. But that performance comes at a price: Wendy's has allocated $100-110 million in capital expenditure for 2025 to scale the system to 500-600 locations. Not every operator has that kind of budget.
When AI Goes Wrong: The Failures Nobody Likes to Talk About
For every Wendy's success story, there is a cautionary tale. McDonald's multiyear IBM partnership ended abruptly in July 2024 after the system generated viral embarrassments -- most infamously, an order for 260 chicken McNuggets and attempts to add bacon to ice cream. The company has since pivoted to Google Cloud, but the IBM experience demonstrated that voice AI in noisy, unpredictable drive-thru environments remains enormously challenging.
Taco Bell's story is equally instructive. After deploying Omilia Voice AI in 650+ stores with Nvidia partnership, the chain faced its own viral backlash -- including a customer who ordered 18,000 cups of water to test the system's limits. Taco Bell's chief digital officer acknowledged even he has mixed experiences with the technology. The chain now advises franchise teams to treat voice AI as a "sometimes tool," best paired with human staff during peak hours.
AI Ordering: Expectations vs. Reality
Myth: AI ordering is plug-and-play -- just turn it on and watch the savings roll in
Reality: McDonald's spent years on its IBM pilot and still pulled the plug. Successful deployments like Wendy's FreshAI require deep integration with POS systems, menu databases, and ongoing tuning.
Myth: AI will replace all drive-thru workers within the next few years
Reality: Current AI handles order-taking but still needs humans for food prep, quality checks, and exception handling. 22% of AI orders still require human intervention according to Intouch Insight data.
Myth: Customers hate talking to robots and will avoid AI-enabled locations
Reality: KFC India saw 67% of customers prefer WhatsApp ordering over phone calls. Intouch Insight found 97% AI satisfaction vs. 91% human. The resistance depends on execution quality, not the concept.
Myth: AI accuracy already matches or beats human order-takers across the board
Reality: Industry-wide AI accuracy sits at 83% vs. human 87%. Wendy's claims 99%, but that is the exception after massive investment, not the rule for typical deployments.
Beyond the Speaker Box: The Chat-Thru Revolution
While drive-thru voice AI dominates the headlines, a quieter revolution is happening on a screen near you. Chat-based ordering through WhatsApp, SMS, and in-app messaging is growing rapidly in global markets -- and arguably delivering better results with lower risk. KFC India processed over 115,000 WhatsApp orders in its first six months, with 67% of users saying they preferred ordering via chat over calling. OlaClick, a platform powering WhatsApp-based ordering for restaurants, now processes over 1.3 million orders monthly across approximately 50,000 active restaurants.
The advantage is structural. Chat-based ordering eliminates the biggest problem plaguing drive-thru voice AI: the noisy, unpredictable audio environment. Text is unambiguous. Customers can visually review their order before confirming. There are no engine sounds, wind, or multiple passengers speaking at once. And it scales to any channel -- WhatsApp, Instagram DMs, SMS, website chat -- without requiring specialized hardware at each location.
What a Chat-Thru Order Looks Like
Hi, I want to order for delivery
Welcome! Here is our menu. What would you like to order today?
2 combo meals #3 and a large fries
Got it! 2x Combo #3 and 1x Large Fries. Your total is $24.50. Would you like to confirm and pay?
Yes, confirm
Order confirmed! Estimated delivery: 25-30 minutes. You will receive a tracking link shortly. Thank you!
Industry analysts project that chatbots will save businesses approximately $8 billion per year by 2026. And 69% of consumers say they prefer using chatbots for quick communication with brands. The Chat-Thru is not a downgrade from the drive-thru -- for many customers, it is an upgrade.
The Numbers Behind the Shift
Fast-Food AI Deployments: By the Numbers
| Brand | AI Partner | Stores | Key Metric | Status |
|---|---|---|---|---|
| Wendy's FreshAI | Google Cloud | 160+ | 99% accuracy, -22s wait | Scaling to 500-600 by end 2025 |
| Taco Bell | Omilia / Nvidia | 650+ | Now a 'sometimes tool' | Active with caveats |
| McDonald's | Google Cloud | 43,000 (planned) | Accuracy Scales in 12 markets | Major tech overhaul underway |
| Burger King ('Patty') | OpenAI | 500 (pilot) | Monitors drive-thru convos | Nationwide by end 2026 |
| Papa John's | Google Gemini | Systemwide (planned) | First Google Food Ordering agent | Rollout by end 2026 |
| KFC India | WhatsApp Bot | National | 115K+ orders in 6 months | 67% customer preference |
The Labor Equation: Why Operators Cannot Wait
Behind every AI deployment decision is a workforce reality that is hard to ignore. The restaurant industry faces 79.6% annual employee turnover, with 45% of operators reporting persistent understaffing. The National Restaurant Association reported a net loss of 25,500 restaurant jobs in Q1 2025 alone. Finding and retaining staff to stand at a drive-thru window is becoming structurally harder every year.
Meanwhile, 92% of operators reported rising labor costs over the past 12 months, and 68% are actively seeking labor-focused technology solutions. AI ordering is not replacing a thriving workforce -- it is filling gaps that operators increasingly cannot fill with human hiring alone.
The Drive-Thru Staffing Challenge
How AI ordering changes the labor equation
Traditional Staffing
79.6% annual turnover
Constant hiring and retraining cycle that drains management time and budgets
Rising labor costs
92% of operators report year-over-year increases with no relief in sight
Understaffed peak hours
Drive-thru speed and accuracy suffer when key positions go unfilled
AI-Assisted Staffing
AI handles order-taking
Frees existing staff for food preparation, quality assurance, and customer hospitality
Consistent performance
No call-outs, no burnout, no turnover on the AI ordering layer
Smarter hiring focus
Invest recruiting budgets in food quality and hospitality roles instead
What Smart Operators Should Do Now
Action Plan
5 Steps to Prepare for the Chat-Thru Future
Whether you run 1 location or 1,000, start here
Audit your current ordering channels
Map every way customers reach you: phone, drive-thru, walk-in, app, WhatsApp, social DMs. Identify where you lose the most orders and where response times are slowest.
Start with text, not voice
Chat-based AI (WhatsApp, SMS, web chat) is more mature, more accurate, and lower risk than voice AI. You do not need specialized hardware, and customers can review orders visually before confirming.
Pick a single high-volume channel first
Do not try to automate everything simultaneously. Choose the channel with the highest message volume and the most missed or delayed orders. Win there first, then expand.
Measure before and after rigorously
Track order volume, average ticket size, response time, and accuracy for at least 30 days before deployment. Run the same metrics 30 days after. Let data drive your decisions.
Scale based on data, not hype
If AI ordering improves your metrics on one channel, expand methodically to the next. The chains that failed -- like McDonald's with IBM -- tried to go too big, too fast, without enough iteration.
See It In Action
Ready to Bring Chat-Thru to Your Restaurant?
Ordering.co helps restaurants deploy AI-powered ordering on WhatsApp and chat channels in days, not months. No hardware required, no complexity.
What Comes Next: The $49 Billion Opportunity
The voice AI restaurant market alone is projected to grow from $10 billion to $49 billion by 2029. But the real story is convergence. The future is not voice AI or chat AI -- it is a unified conversational commerce layer where the channel does not matter. A customer might start an order on WhatsApp, modify it through a voice call, and confirm via the restaurant's app. The AI remembers context across every touchpoint.
Papa John's deployment of Google Gemini-powered ordering is a preview of this future: one AI brain powering voice ordering, text ordering, group orders, and intelligent deal recommendations -- all through a single system. Industry observers are already calling 2026 "the year of the AI-driven restaurant."
The operators who win will be those who start now -- not with the most expensive voice AI deployment, but with the channel their customers are already using. For most restaurants globally, that channel is messaging. The drive-thru is not disappearing. But the Chat-Thru is arriving faster than anyone expected.
Frequently Asked Questions
Common questions about AI ordering in fast food
How accurate is AI ordering in fast-food drive-thrus?
Industry-wide AI order accuracy averages 83% compared to 87% for human order-takers, according to Intouch Insight 2025 data. Wendy's FreshAI claims 99% accuracy across its 160+ deployed locations, but this remains an outlier requiring significant investment. Accuracy improves as systems learn from each restaurant's specific menu and customer ordering patterns.
Why did McDonald's end its AI drive-thru partnership with IBM?
McDonald's terminated its Automated Order Taking pilot with IBM in July 2024 after persistent order errors. The most notable incident involved a system-generated order for 260 chicken McNuggets that went viral on social media. The company has since pivoted to a partnership with Google Cloud for a broader technology overhaul across its 43,000 restaurants.
What is chat-based ordering and how is it different from voice AI?
Chat-based ordering uses messaging platforms like WhatsApp, SMS, or web chat instead of spoken conversation at a drive-thru speaker. It eliminates audio noise problems that plague drive-thru voice AI and lets customers visually review their orders before confirming. KFC India processed over 115,000 WhatsApp orders in its first six months with 67% customer preference.
How much does it cost to implement AI ordering in a restaurant?
Costs vary significantly by scale and technology approach. Wendy's allocated $100-110 million in capital expenditure for 2025 to scale FreshAI to 500-600 locations. For independent and mid-size restaurants, chat-based AI solutions through platforms like Ordering.co can be deployed in days without hardware investment at a fraction of the cost.
Will AI ordering replace restaurant workers?
AI ordering automates the order-taking function but does not replace the full restaurant workforce. Staff are redeployed to food preparation, quality assurance, and customer hospitality. Given the industry's 79.6% annual turnover rate and persistent understaffing, AI addresses a labor shortage rather than displacing available workers.
Which fast-food chain is leading in AI ordering technology?
Wendy's FreshAI currently leads in deployed performance with 99% reported accuracy and 22-second wait time reduction across 160+ locations. McDonald's Google Cloud partnership covers the largest footprint at 43,000 restaurants. Taco Bell has the highest store count with 650+ deployments, though it has scaled back to a hybrid human-AI approach.
Does AI ordering change what customers order?
A February 2026 Penn State University study published in the International Journal of Hospitality Management found that AI ordering increased indulgent food choices by approximately 40%. Researchers found that interacting with AI requires more cognitive effort, leaving customers less able to resist indulgent options compared to ordering from a human.
How large is the AI market in quick-service restaurants?
The AI market in quick-service restaurants was valued at $915.3 million in 2024 and is projected to reach $12 billion by 2034, representing a 29.4% compound annual growth rate. The voice AI restaurant market specifically is expected to grow from $10 billion to $49 billion by 2029, driven by labor shortages and consumer demand for faster service.
Your Customers Are Already Chatting. Are You Listening?
Deploy AI-powered WhatsApp ordering for your restaurant. No hardware, no complexity, real results.