Engineered for Delivery

Dan McFadden - Co-founder/CTO

Oct 24th, 2019

First Note

That was the single note I scribbled on my notepad after a 2 hour meeting with my soon to be co-founder, Chris Baggott. That note would serve as our north star as we created a superior delivery company.

Chris had been thinking about starting a food delivery business to fix the brokenness in the current third-party models. At that time GrubHub was the big player and they had gone public. He was intrigued to see their stock price going up while the customer, restaurant and driver experience was going down. “Let’s eliminate the third party friction by eliminating the third party.”

Chris asked me if I’d be interested in joining him. He’d raise some money to get us started and I’d build it. At that time I was working for Oracle after our previous company was acquired by them. The idea of joining Chris again was super appealing to me, I have always appreciated his vision. He seems to live 5 to 10 years in the future. The trick is capturing that vision and executing on it. That night I talked it over with my wife and we decided now was a perfect time in my life to take a risk. We had no kids at the time and she had a stable job. So, I let Chis know I was in and we got to work.

Building a ClusterTruck

Mission #1: Find a building that could host our kitchen and HQ. That happened pretty quickly and before we knew it hammers were swinging at 729 N Pennsylvania. It was the perfect building. Main floor for the kitchen, the second floor for our office and an easy exit onto Penn for our drivers. Now I was racing the clock. I told myself we are never going to hold up the launch because the software is not ready. Our deadline is when Pad Thai’s start getting test cooked in that kitchen. It was fun to race against and honestly the motivation we needed to pull this off. I remember walking in there every other day and checking their progress and comparing it to ours.

🔨 Hammering all the things

Now it was time to assemble a small team, design all this stuff and build it. I had banked on being able to pull a few old co-workers out of Oracle and utilize some off the shelf software to power the kitchen and the delivery. This meant we really just needed to focus on the customer ordering part… Easy stuff.

Early Team Picture
Early pic of the team around September 2015

Time to roll our own 😬

Well, after a solid week of demoing POS systems and talking to sales people (ugh), I discovered there was no POS system in world that is 1) not terrible, and 2) going to meet our needs.

So I told my guys, “Surprise! We’re rolling our own kitchen software.” They were not amused. But, we quickly got on the whiteboard and started hashing it out.

Early Kitchen Prototype
Early Kitchen Prototype

Drivers First

A month later it was time to figure out the driver part. Great, now we can cook the food, but how do we get it to the customer? Again, I had planned on pressing the easy button and using Uber to do our deliveries. After digging in it became clear that it would simply not work. The main problem was we could not uphold the mission: “Don’t cook the food until the driver is the correct distance away.” So… we started building out a driver platform. All of these roadblocks were huge hits and very discouraging at the time, but they turned out to be the biggest blessings. One of the things I’m most proud of in my career is the system we were forced to invent to manage our drivers. It’s elegantly simple, super efficient and allows them to make great money, win win win.

March 2016: Order #1 Ships 🎉🍔📈

We shipped the first order out of 729 end-to-end completely on our own software. This gave us a HUGE advantage and even more satisfaction. We had complete freedom to tune and customize each and every piece. It was also pretty damn cool to say we pulled this off with a team of ~3 guys in < 8 months.

Order 1
Order #1 to my team member Andrew (missing in pic below)

Post launch we painstakingly focused on fixing, improving and iterating. The first couple months were very very rough. Everything broke literally every day. I remember thinking “we are never going to get this thing stable.” 50 orders used to break us… Fast forward 3 years and we do 400 orders at lunch without flinching. Remarkably, all on the same “hardware” (the kitchen) but, with much better software. It’s an amazing thing to see in action. Our Indy kitchen has to be the busiest kitchen in the world during lunch rush.

OK, cool…but is it done?

Team Picture
Our current team. From the left: John, Tylor, Dan (Me), Daniel, Brian, Chris and Arda, Andrew (not pictured), Callie (not pictured)

A question I often get is: “What do you guys do all day now that the software is done?” The person who asks me this most is my wife 🤣 She’s a hygienist, so I respond to her “I flossed and brushed my teeth yesterday… am I’m done?” And, she rolls her eyes and me and walks away.

I guess to somebody who is not a software developer the question is valid, I guess. Software is like anything else, it requires maintenance and improvements. Without that it rots and dies. We’re relentless about improving our product and customer experience. We strive to continue to widen the moat around us. Sure, VC backed third-parties have way more cash and engineers than we do, but we have a better model. And, You can’t buy or engineer your way out of a broken model. Eventually the VC cash runs out and you gotta make money.

I write all this for a few reasons:

0. Thank my team for their hard work and dedication
1. Personally reflect on the past 4 years of my life
2. Give our fans and customers a different perspective on ClusterTruck
3. Maybe inspire somebody who is thinking about starting something new

Starting new things is hard, very hard. I honestly thought we would never even get ClusterTruck launched, it was just so complicated. We recently shipped our 1,000,000th order! I’ve learned that if you surround yourself with great people and break big problems down to the “simplest thing that can possibly work,” anything is possible.