Building software got too easy, so I started a courier company
I've been building software for other people for over 10 years. Startups, client projects, agency work. For the last few years I've run Solveita, a one-man dev agency in Malta, and I still do.
This year I started something that doesn't look like a software business at all. It's a courier company called Kwip, next-day parcel delivery for small online shops in Malta.
Why a dev would do this
Because building got cheap. Something that took me a week now takes a day with Claude Code. That's great for me, but it takes everyone else a day too. Whatever I ship, someone can copy before I've finished writing the launch post. I don't think software is dead, but as a solo founder with no distribution, pure software felt like a worse bet every month.
A courier company works the other way round. You need a van, a base, insurance, a postal authorisation from the regulator, and someone has to physically show up every single day. Nobody clones that over a weekend. The moat is boring and expensive, and that's exactly why I like it.
Where the idea came from
My wife runs a beauty salon and sells products online. Every courier she tried was built for big shippers: morning cutoffs, CSV manifests, monthly minimums. When a delivery failed, the customer blamed her, not the courier. I spoke to a bunch of small Maltese merchants and heard the same story. Their options were deliver it themselves, lose the sale, or stop selling online.
One stat sealed it for me. According to Eurostat, Maltese online shoppers buy more often than almost anyone in the EU (33% made 10+ purchases in three months), yet only about 21% of small Maltese businesses sold online in 2024. The customers are ahead of the shops.
What Kwip does
- No minimum volume, no contract, no cutoff time
- Merchants book a pickup when the parcel is ready, from slots based on real van capacity
- Delivery by next day from pickup, Monday to Saturday, Gozo on Saturdays
- From €3.99 + VAT per drop, and every extra is a flat €1
The software is where my background actually counts. There's no dispatcher role in the system at all. Pickup slots are calculated from the van's route and the driver's calendar, the system decides what travels together, and recipients get a time window that tightens through the day instead of "sometime between 9 and 6". They can reschedule over WhatsApp without anyone calling a depot.
Stack, for anyone curious: Laravel, Next.js, PostgresQL, Stripe, and WhatsApp API.
Two questions for you
- Anyone here running a physical business with a software edge? What broke first?
- Later on I could sell the dispatch software to other small couriers. Would you start that early, or stay focused on the courier until it runs itself?
You can see what we're building at kwip.mt, and I'll be posting updates on the Kwip product page here.
Agree on the moat, and it changes where your marketing lives too. For a local service, the search that matters is a shop owner typing "next day courier Malta", or asking ChatGPT the same thing. Neither rewards clever software. They reward a complete Google Business Profile, reviews from real shops, and a listing in the two or three local directories everyone already uses.
It's also the cheapest thing to do while the first shops come in by hand. Ask every early customer for a short review that names the service and the town, because those are the exact words people search.
We're the opposite case at UtilitySEO: pure software at DA 3 with about four search visits a month, so I slightly envy a business where one van and good reviews can win the local result.
Is Kwip on Google Business Profile yet?
The software may solve a real operational problem, but demand from other couriers is a separate assumption. What would you need to see before treating it as a standalone product?
This is the most honest take on 2026 I've read here. When anyone can ship the app in a weekend, the moat moves to the stuff that can't be prompted: vans, routes, relationships with shop owners. How are you getting your first shops, walking in and asking, or reaching them online? I'd guess for a local service the in-person ask wins every time.
The slot logic you described hangs on one van and one calendar. A second operator with a depot, or a hard cutoff, turns "no dispatcher" into a list of exceptions, and that is the product you would be selling. I would keep the software inside Kwip until someone else asks for that same constraint.