Showcase · Delivery
The software could not hold the process. So the process was being thrown away to keep the software.
A spring water delivery business in Kelowna, BC. One truck, one owner, and every customer a person the owner knows by name. The client needed to run customers, routes, money and messages from the road, while delivering.
Press it yourself. It is live. Book a walkthrough
The real app with made-up customers. Nothing is sent and nothing is charged. It takes about five seconds to set up your own copy.
Three delivery days a week. About 50 doors a day, 150 a week, and every one of them is on the app.
Missed deliveries, double routes, missed payments and complaints: down to almost zero.
The client
A solo operator, just starting, with a business that runs on standing promises.
The client fills the jugs, drives the run and knocks on every door. The business is built on small arrangements made in person: the customer who pays a different price agreed at the door, the person who is only home after three, the neighbour who sold the client jugs and is being paid back in water.
None of that is untidy. It is the business. Any software had to hold all of it.
The problem, in the client’s words
Two months on a big platform, and still nowhere.
- Every week I texted every customer one at a time. Half did not answer, so I texted again. I kept who owed what in my head and chased it by hand. I planned the route on the kitchen table.
- I bought software. Onboarding went on for nearly two months. Round after round. It was built for a generic delivery business, and mine is not one.
- I found myself redesigning how I actually work to fit the tool. Route logic, payment terms, the promises I had made to people at their doors, all had to be bent or dropped.
- I needed to run the whole thing from the truck. Customers, routes, money, messages. While I am out delivering.
How we got there
We learned the software that had failed them before we wrote a line of our own.
The client wrote down how the business really runs. Not a feature list. The actual working day, including the parts that are not tidy. We built to that.
-
01
Listen.
The client called, frustrated. We asked for the real day: who gets a text, who gets an email, who pays cash, who is only home after three.
-
02
Learn the old tool.
We took the big platform’s own training course ourselves, start to finish, so we understood exactly what the client had been fighting for two months.
-
03
Build a demo.
A working version, three doors: the customer, the truck and the office. The client pressed everything and said: I like this, I do not like this.
-
04
Three or four rounds.
We redid the code each time. Every round moved the app closer to how the business already ran, never the other way.
-
05
Run it beside the old way.
Three deliveries in parallel, with ten trusted customers, to shake the bugs out before anybody else saw it.
-
06
Keep going.
Retained. One update a week as needed. Sometimes it is a whole feature. Sometimes it is moving a button to the other side of the phone so it can be pressed one-handed.
The route optimiser came later. The client knew the city by heart, then asked what it would cost. It existed within 24 hours, ran once with a few bugs, and was fixed for that Thursday’s run.
What we built
One app with three doors: the customer, the truck and the office.
It asks every customer for their order on its own, books the standing orders, builds the run in driving order from the spring, sends one invoice per customer with one link to pay, and draws down prepaid packs at the door. The client drives and talks to people. The app does the rest.
It started as a delivery app. It has grown into the whole operating system for the business, and it is still growing, one week at a time.
The small things, coded in
This customer gets a text. That one gets an email. That one pays cash.
The point of the job was never the big features. It was every small, real rule of the business, held by the software instead of by the owner’s memory.
-
Ordering
Customers order, skip a week, or set a standing order from their own app. They are never asked for anything the business already knows.
-
The run
Every delivery day is built and sequenced from the fill point, on real map data. The customer who only wants water after three gets the truck at three.
-
Money
Invoices, prepaid packs, cash at the door, part payments and past due, all in one ledger. One customer, one balance, one link to pay.
-
At the door
A photo of the delivered water for proof. Delivered, cash taken, balance moved. The customer’s app updates before the truck has left the driveway.
-
Never-Out
Out of water before the next run? One tap. The owner gets it at once and it lands on the day’s route.
-
Referrals
Every customer carries their own link. The app tracks the credit and applies it to both sides. Almost 40 percent of new customers now come this way.
The result
The owner’s attention is back on the two things only the owner can do: driving, and talking to people.
Live for a little over ninety days. The weekly asking, the reminders, the invoicing, the route, the balances and the follow-ups run without anyone touching them. Missed deliveries, double routes, missed payments and customer complaints have dropped to almost zero. The client has evenings again.
- 150doors a week, all on the app
- 3delivery days a week
- 40%of new customers by referral
- 24 hfrom “can it plan the route?” to a route optimiser
Press it yourself
The same app, with made-up customers.
Be the customer, then the office, then the truck. Order water, press Never-Out, read the morning, run the route on the map and mark a door delivered. A ten-step tour walks you through it. Nothing is sent and nothing is charged.
Opens in a new tab. About five seconds to set up your own copy. Press Start over and it all goes back.
It has grown into a small ERP for a business with one truck. That was never the pitch. It is what happens when you model the real process instead of buying someone else’s.
This is one job. Yours will not look like it.
Anyone can generate code. The hard part is understanding a business well enough to know which parts must not be compromised, and building around those. Tell us the worst job in your week.
Book a walkthrough See the other showcases
Free. Thirty minutes. You keep the map.