Pizza Huts $100MIL AI failure 🥶🍕

@kylercheatham

24.5K followers · usually 5.3K views a post

Views

25.6K

Likes

2.1K

Comments

195

Selling · SaaS › Other SaaS

Dragon Tail AI

Why it worked

The video uses a relatable and cautionary tale about a major company's AI failure, which resonates with viewers interested in business, technology, and operational efficiency. The specific details of the lawsuit and the financial implications make the story compelling and shareable.

Summary

The video discusses Pizza Hut's $100 million AI failure, highlighting how a mandatory rollout of the Dragon Tail AI delivery optimization tool led to operational breakdowns and decreased performance. The failure stemmed from a lack of governance, poor implementation, and ignoring real-world data and franchisee feedback, ultimately costing the company millions.

Structure

1Introduction to Pizza Hut's AI failure and its cost.
2Explanation of Yum Brands' acquisition of Dragon Tail AI and its purpose.
3Details of the rollout and the negative impact on delivery times and operations.
4The lawsuit filed by Chaac Foods against Pizza Hut for forcing the AI implementation.
5Key takeaways: mandatory rollouts without adoption/governance fail, and ignoring data leads to costly mistakes.

Product placement

Yum Brands (parent company of Pizza Hut) ACTS: The video discusses Yum Brands' acquisition of Dragon Tail AI and its implementation. Removing the name would not change what happens in the video. Pizza Hut ACTS: The video discusses Pizza Hut's AI failure and its impact on operations. Removing the name would not change what happens in the video. Dragon Tail AI ACTS: The video discusses Dragon Tail AI's role in delivery optimization and its implementation by Pizza Hut. Removing the name would not change what happens in the video. Chaac Foods ACTS: The video mentions Chaac Foods as a company that acquired Pizza Hut chains and later sued Pizza Hut. Removing the name would not change what happens in the video. DoorDash ACTS: The video mentions DoorDash as a gig economy platform that drivers use and that Pizza Hut's AI interacts with. Removing the name would not change what happens in the video.

Original caption

I don’t know about you, but I draw the line at cold Pizza ⛔🍕##aifail##corporatelife##digitaltransformation##businesssystems##techtok

More from @kylercheatham

Their other posts in the index, biggest breakout first.

On-screen text

Pizza Huts $100MIL AI failure 🥶🍕 WHAT CAN COLD PIZZA TEACH US ABOUT AI FAILURE? GLAD YOU ASKED. BECAUSE IT CAN ACTUALLY COST YOU $100 MILLION PIZZA HUT. OF THOSE CLASSIC EXAMPLES OF WHERE CORPORATE SAYS AI WILL FIX EVERYTHING AND IT ACTUALLY ENDS UP BREAKING EVERYTHING AND COSTING PEOPLE A BUNCH OF MONEY. HAS A COUPLE WE'VE GOT YUM BRANDS, WHICH IS OF PIZZA HUT, RIGHT? WE'VE GOT CHAAC FOODS, THAT'S OUT OF NEW ENGLAND, OR WE, THEY PROBABLY SAY SHOCK FOODS, UM, THAT IS FOUNDED IN 2006, BUT IN 2020, IT ACQUIRES OVER 100 DIFFERENT PIZZA HUT CHAINS THROUGHOUT NEW ENGLAND, NEW JERSEY, UM, NEW YORK, PHILADELPHIA, MARYLAND, DC, ALL THOSE AREAS. NOW, IN 2021, YUM BRAND SAYS, THE PIZZA HUT ACQUIRES DRAGON TAIL AI TO STAY, IN THE AI GAME. AND REALLY, WHAT THAT IS IS A DELIVERY OPTIMIZATION TOOL. NOW, IN 2024, YUM BRAND SAYS, WE'RE ROLLING OUT DRAGONTAIL AI IN ALL OF OUR NEW YORK STORES WITH CHAAC FOODS IS INCLUDED IN THAT. AND NOW IT'S IMPORTANT TO UNDERSTAND THIS AI ROLLOUT, CHAAC FOOD LOCATIONS ARE KILLING IT, ESPECIALLY IN NEW THEY'RE JUST. OVERALL OVER 10 PER 10% YEAR OVER YEAR, AND THEY DELIVER PIZZA WITHIN 30 MINUTES 90% OF THIS WAS NOT LIKE A LAGGERED FRANCHISEE NEEDING RESCUE. HIGH PERFORMING, HIGHLY PROFITABLE AREA. THE BUSINESS THAT DIDN'T NEED TO AI IN 2025 AFTER DRAGONTAIL AI WAS IMPLEMENTED, ALL OF A SUDDEN THEIR PERFORMANCE STARTS TO PLUMMET. DELIVERY TIMES ARE SCRATCHING FROM WHAT USED TO BE 30 MINUTES 90% TO OVER 45 MINUTES MORE THAN HAlf OF THE TIME NOW AT AND THE FRANCHISEE ALLEGES THAT THERE HUGE OPERATIONAL BREAKDOWN AND PIZZAS SITTING ON THE RACK FOR 15 TO 25 MINUTES. KNOW THAT TRICKLES DOWN TO CUSTOMER SATISFACTION, BECAUSE YOU'RE REALLY ONLY ORDERING PIZZA HUT IF YOU AND A LOT SO IT'S IMPORTANT THAT DRAGON TAIL. I THINK I'VE CALLED IT DRAGONFLY HALF THE TIME, THAT. BUT DRAGON TAIL AI, THE, THE POINT OF IT IS ORDER TO DELIVER MANAGEMENT PLATFORM, WHICH MEANS IT OPTIMIZES, LIKE, THE FULL LIFE CYCLE OF THE ACTUAL PIZZA FROM THE KITCHEN RIGHT TO THE DRIVER DISPATCH. TO MAKE IT FASTER SO YOU CAN GET MORE IT'S HIGH VOLUME PLAY HERE. CORPORATE PIZZA HUT BRAIN OF THE KITCHEN, OPTIMIZE THIS FOR DELIVERY DRIVERS. BUT WHAT THEY DIDN'T REALLY COMPREHEND IS THE GAMING OF THE SYSTEM YOU GUYS ARE GIVING TOO MUCH VISIBILITY INTO KITCHEN OPERATIONS TO THESE GATE DRIVERS LIKE DOORDASH, RIGHT? AND THEY'RE STAYING IN THE STORE UP TO 15 MINUTES BECAUSE THEY'RE TRYING TO BASH THESE ORDERS. THE RACK TIME RIGHT FROM. PIZZA COMES OUT INTO THE COOLING WRAP, WENT FROM, ON AVERAGE OF FIVE MINUTES WHEN THE PICK UP THE COME BACK, PICK UP THE PIZZA, DELIVER IT, TO ABOUT 20 MINUTES, BECAUSE NOW THEY'RE THIS PIZZA'S ALMOST ON THE RACK, GONNA WAIT 10 MORE MINUTES, GET FIVE PIZZAS, AND THEN DELIVER THEM. THERE WAS NO GOVERNANCE AROUND WHERE DRAGON TAIL STOPS. PLATFORMS LIKE DOORDASH STARTED. SO THESE GIG ECONOMY DRIVERS COULD SEE WHICH TYPE OF ORDERS WERE CASH OR WHICH TYPE OF ORDERS TIPPED WHAT. AND THEY PRIORITIZE BASED ON THEIR OWN INCENTIVES, KNOW, ANY HUMAN WORKING THOSE JOBS WOULD. A COMPLETE LACK OF STANDARDIZATION IN KIND OF CHERRY PACK PICKING, BECAUSE THEY STACKED TWO ALGORITHMS ON TOP OF EACH OTHER WITH NO GOVERNANCE IN BETWEEN THEM. LEADS TO MAY 2026, WHEN CHAAC FOODS TAKES OUT A TEXAS BUSINESS COURT AGAINST PIZZA HUT, SAYING, YOU FORCED US TO UTILIZE THIS PLATFORM, WHICH WE NEVER ASKED FOR, RIGHT? UNLESS YOU TANKED OUR MARGIN AND YOU LOST US MONEY. SO THEY SUED THEM FOR $100 MILLION. SO WHAT ARE THAT WE CAN WHICH IS THE REASON WHY WE OF CASE STUDIES? WELL, ONE, A MANDATORY ROLLOUT NEVER WORKS THE WAY YOU THINK IT DOES, ESPECIALLY WITHOUT ANY SORT OF ADOPTION HAVE THE CENTRALIZED AI PLATFORM INTO CONSIDERATION ANY SORT OF LOCAL OPERATIONS, OF KILL SWITCHES, THAT THEY COULD OFF OR STOPPED IT WHEN THEY STARTED TO SEE THEIR KPI'S TANK. SOMETHING WE SEE ALL THE TIME WITH PLANTS THROUGHOUT THE WORLD. OF NOT EVEN THOUGH YOU'RE UNDER NO TWO DIFFERENT MANUFACTURING PLANTS ARE THE SAME. A HUGE FAILURE ON THEIR PART. THIS IMPLEMENTATION OF DRAGON TAIL LIKE, A SILVER BULLET. LIKE, HERE, YOU'RE WELCOME. GONNA REVOLUTIONIZE YOUR KITCHEN OPERATIONS. WITHOUT UNDERSTANDING OR ANY OTHER SYSTEM IMPLEMENTATION. WITH DISCIPLINE, TRAINING. RIGHT. THEY JUST SAID, ROLLOUT, IT'S GONNA THERE'S GONNA BE NO ISSUES WITH DIDN'T SUPPORT THESE LOCATIONS IN ACTUALLY TRAINING ON THE, YOU KNOW, LOCAL OPERATIONS AND INTEGRATION WITH DOORDASH. ONE OF THE BIGGEST ONES IS THEIR KPIS WERE TREATED MORE LIKE DASHBOARDS. WHEN, YOU KNOW, THOSE KPIS SHOULD AND YOU SHOULD TURN IT OFF. RIGHT. SEEING ALL OF THIS DATA AROUND, WORKING, AND YOU'RE UNABLE TO INFLUENCE THAT SYSTEM BECAUSE IT'S THAT'S A HUGE ISSUE. IT'S NOT A TRANSFORMATION. YOU HAVE LIKE A RUNAWAY EXPERIMENT THAT YOU. PILOT TO BE ABLE TO KNOW WORK OR NOT. IF YOU IGNORE YOUR BUSINESS MODEL JUST TO PRIORITIZE AI, I PROMISE YOU, IT WILL BREAK ON AN EDGE AI PRODUCTS, EVEN IF THEY ARE VERTICAL BASE. THIS ONE, THAT TREAT OPERATIONS LIKE SOME INTERCHANGEABLE ABSTRACTIONS, IS ALWAYS, ALWAYS GONNA BE TRIPPED UP ON EDGE EDGE CASES. ESPECIALLY WHEN THOSE EDGE CASES ARE YOUR BIGGEST FRANCHISEES AND MONEYMAKERS. AND THIS UNFORTUNATELY OFTEN HAPPENS WITH FRANCHISEE MODELS BECAUSE THEY TREAT SORT OF TEST ENVIRONMENTS INSTEAD OF LIKE VIABLE PARTNERS, ESPECIALLY HIGH PERFORMING ONES. BUT ULTIMATELY, AI IS JUST A TOOL. IF YOU REFUSE TO LISTEN TO THE REAL, LIVING, BREATHING DATA THAT'S YOUR CUSTOMER EXPERIENCE, IT WILL FAIL AND COST YOU HUNDREDS OF MILLIONS OF DOLLARS.

Transcript

What can cold pizza teach us about AI failure? Glad you asked. Because it can actually cost you $100 million. Pizza Hut. Of those classic examples of where corporate says AI will fix everything, and it actually ends up breaking everything and costing people a bunch of money. Has a couple. We've got Yum Brands, which is of Pizza Hut, right? We've got Chaac Foods, that's out of New England, or we, they probably say shock foods, um, that is founded in 2006, but in 2020, it acquires over 100 different Pizza Hut chains throughout New England, New Jersey, um, New York, Philadelphia, Maryland, DC, all those areas. Now, in 2021, Yum Brand says, the Pizza Hut acquires Dragon Tail AI to stay, in the AI game. And really, what that is is a delivery optimization tool. Now, in 2024, Yum Brand says, we're rolling out Dragon Tail AI in all of our New York stores with Chaac Foods is included in that. And now it's important to understand this AI rollout, Chaac Food locations are killing it, especially in New. They're just. Overall, over 10 per 10% year over year, and they deliver pizza within 30 minutes, 90% of this was not like a laggered franchisee needing rescue. High performing, highly profitable area. The business that didn't need to AI in 2025 after Dragon Tail AI was implemented, all of a sudden their performance starts to plummet. Delivery times are scratching from what used to be 30 minutes, 90% to over 45 minutes more than half of the time now at and the franchisee alleges that there huge operational breakdown and pizzas sitting on the rack for 15 to 25 minutes. Know that trickles down to customer satisfaction, because you're really only ordering Pizza Hut if you and a lot. So it's important that Dragon Tail. I think I've called it Dragon Fly half the time, that. But Dragon Tail AI, the, the point of it is order to deliver management platform, which means it optimizes, like, the full life cycle of the actual pizza from the kitchen right to the driver dispatch. To make it faster so you can get more. It's high volume play here. Corporate Pizza Hut brain of the kitchen, optimize this for delivery drivers. But what they didn't really comprehend is the gaming of the system. You guys are giving too much visibility into kitchen operations to these gate drivers like DoorDash, right? And they're staying in the store up to 15 minutes because they're trying to bash these orders. The rack time right from. Pizza comes out into the cooling wrap, went from, on average of five minutes when the pick up the come back, pick up the pizza, deliver it, to about 20 minutes, because now they're this pizza's almost on the rack, gonna wait 10 more minutes, get five pizzas, and then deliver them. There was no governance around where Dragon Tail stops. Platforms like DoorDash started. So these gig economy drivers could see which type of orders were cash or which type of orders tipped what. And they prioritize based on their own incentives, know, any human working those jobs would. A complete lack of standardization in kind of cherry pack picking, because they stacked two algorithms on top of each other with no governance in between them. Leads to May 2026, when Chaac Foods takes out a Texas business court against Pizza Hut, saying, you forced us to utilize this platform, which we never asked for, right? Unless you tanked our margin and you lost us money. So they sued them for $100 million. So what are that we can which is the reason why we of case studies? Well, one, a mandatory rollout never works the way you think it does, especially without any sort of adoption have the centralized AI platform into consideration any sort of local operations, of kill switches, that they could off or stopped it when they started to see their KPI's tank. Something we see all the time with plants throughout the world. Of not even though you're under no two different manufacturing plants are the same. A huge failure on their part. This implementation of Dragon Tail like, a silver bullet. Like, here, you're welcome. Gonna revolutionize your kitchen operations. Without understanding or any other system implementation. With discipline, training. Right. They just said, rollout, it's gonna there's gonna be no issues with didn't support these locations in actually training on the, you know, local operations and integration with DoorDash. One of the biggest ones is their KPIs were treated more like dashboards. When, you know, those KPIs should and you should turn it off. Right. Seeing all of this data around, working, and you're unable to influence that system because it's that's a huge issue. It's not a transformation. You have like a runaway experiment that you. Pilot to be able to know work or not. If you ignore your business model just to prioritize AI, I promise you, it will break on an edge AI products, even if they are vertical base. This one, that treat operations like some interchangeable abstractions, is always, always gonna be tripped up on edge edge cases. Especially when those edge cases are your biggest franchisees and moneymakers. And this unfortunately often happens with franchisee models because they treat sort of test environments instead of like viable partners, especially high performing ones. But ultimately, AI is just a tool. If you refuse to listen to the real, living, breathing data that's your customer experience, it will fail and cost you hundreds of millions of dollars.