Sometimes, clever people think they’ve got all the answers. They rush into building a product, in a garage and hidden from the world. They forget to talk to customers and build with 1,000 blind assumptions. Then once it’s released, they expect everyone to come running and pay their last dollar to get their hands on this magical product.
Except often, it doesn’t work as planned. These clever people find that the world doesn’t need what they’ve built. Often, no matter how shiny or wifi-connected it is, it remains unnecessary.
I’m looking at you, Juicero, the company that raised $120m to make robots that squeeze oranges with 4 tons of force. The company that sold these robot juicers for $700 per machine. The company that was founded in order to make the “greatest impact on humanity.” With juice.
Venture Capitalists love these “opportunities.”
How I feel about $700 internet-connected juicers.
The self-driving car industry has a lot of these clever people. Messiah-complex folks. “I-can-change-the-world-because-I-know-the-truth-and-you’re-all-lucky-I’m-here-to-give-it-to-you”, folks.
Putting the customer first: analyze; test; learn; iterate; expand
At Udelv, we’re taking a different approach. We believe in the power of listening to the Voice-of-Customers using our product, not to the Voice-of-God sitting on our shoulders. We believe in putting our work out into the world early and watching the results unfold. We believe that mistakes can be good.
If there is one industry that is unforgiving like none other, it’s automotive. This industry needs to be approached with humility. For every bold Tesla, there are thousands of failed attempts.
This strategy helps us understand our value to our clients and to their customers from all aspects of the customer experience and product fit, all the way down to on-the-ground training and cost structure. We learn about the tools we need to build, and the experiences we need to provide in order to make autonomous delivery vehicles (ADVs) part of the fabric of 21st-century logistics.
At Udelv, we launched our first ADV in January 2018 and started making deliveries, and mistakes, right away.
Throughout our 6,000 deliveries to date, we’ve made many mistakes and learned many lessons. Below, we cover a few of those stories, and what they taught us.
Draeger’s Market was Udelv’s first customer. At the time, we were better at photography than autonomous delivery.
Mistake: No backup plans
Our 2nd ever delivery was to a potential investor. As we arrived, the investor tapped the button on his Udelv app to open the compartment. Nothing happened. 🤦
We ended up prying the compartment open by hand and delivering his groceries along with an apology.
Lesson Learned: Customers expect infallibility
We’ve got a Murphy’s Law poster hanging in our lobby. Anything that can go wrong, will go wrong. A fully formed autonomous delivery solution needs to have fallback plans to avoid these issues.
After this mistake, we doubled down on the operations behind the curtain. We gave our teleoperators the ability to control compartment openings from our offices, ensuring that customers are never stuck waiting for a driverless vehicle to respond. Since this mistake, our team is ready to help with every delivery.
Mistake: Some people don’t want our service.
Our 18th delivery was for a florist. A husband had ordered flowers for his wife. Our ADV arrived at the wife’s office and sent her a text saying “Your flowers have arrived. Come outside and retrieve them.” 5 minutes later, a woman showed up, quite displeased that we make her get her own flowers. We’re still unsure if she was mad at us or at her husband.
Burlingame Laguna Florist taught us a lot about customer expectations.
Lesson Learned: The customers of our customers are our customers, too
We had assumed that our solution can help every market that needed us. Florists often can’t hire enough drivers to make all the deliveries requested of them. “Of course they’re going to love ADVs”, we thought.
But we overestimated our value, and didn’t consider the needs of our clients’ customers. This experience taught us to be a lot more thoughtful about the end-to-end experience.
Mistake: Limiting order retrieval
Our 150th delivery was for a grocery store. We arrived at the customer’s home and sent the ‘come out and retrieve’ message to the phone number on file. Turns out, the woman who placed the order, whose phone number was tied to the order, wasn’t home. Instead, she had told her kids to wait on the porch until a funny orange car pulled up. They had no way to retrieve their goods, so they knocked on the ADV window until we opened the compartment ourselves (using the teleoperations tool we built after mistake #1!)
Lesson Learned: Customers work in teams
The customer placing the order is not always the one receiving it. We built the ability to share order retrieval access (read more) with friends and family. This turned out to be one of our most popular tools, often used by the customers of our clients (lesson #2 coming into play!).
Mistake: Forgetting about the nuts and bolts
Udelv’s Version 2 uses a patented iris system to secure orders of different sizes.
Somewhere around delivery 1,500, we started delivering groceries for Walmart using our Version 2 ADV. All of a sudden, the cargo space we were so proud of failed, and it took a quick look to find the cause: a plastic bag got stuck in our mechanics.
Lesson Learned: There is always more to learn!
All those deliveries we had made, and we never accounted for flowing plastic bags. We’re currently working through a re-design to solve the issue, so customers can expect a more reliable service. More importantly, we learned that every industry presents unique challenges. The best way to handle them all is to put something into the world and watch what happens.
–
We’ve made a ton of mistakes in the past 2 years of autonomous deliveries. We’re proud of these mistakes. We’re excited to talk about them. And we think these mistakes allow us to build a better product, enabling us to remove the driver from a delivery sooner than anyone else.
There are no illusions of grandeur here. No masterminded plan, no secrets hidden in a lab, released after a decade of research just to fall upon deaf ears. We put our work into the world, listen to feedback, and improve.
If you want to learn more about our history of mistakes and improvements, or are curious about how the tools we’ve built work for your industry, email us at info@udelv.com.