It all started about 7 years ago.
My mother got laid off and had to find a job. She already had a large garden and decided to make a living selling vegetables. She invested little capital as the activity was supposed to last only a few years until retirement.
Therefore the production was small. Like, very small. She tried selling at the local market but you need large volumes to make the trip worth it. Clients always leave the last vegetables left in each box, they don't want to take the last few tomatoes. Turns out, those tomatoes were just the only ones she had at first.
So she decided to try another sales channel : she would sell vegetables by email. Each week, she'd send a list of all the produces that were available in the garden. The clients (mostly friends and acquaintances at first) were free to order by simply replying to the email with what they wanted. At the end of the week, she'd prepare each order and clients would come to the farm to pick up their order. They would pay directly on location.
Preparing orders for clients is a lot of extra work, taking orders is also quite a process, and the whole thing was a bit inefficient, but she did manage to sell vegetables.
I was a student at the time at university and needed a project for a university class. With a few friends we decided to start a project that'd automate most of the process of offering products and taking orders.
The business model was to create a great app for my mother, and sell it to other farmers later on as a SAAS.
The first iteration was based on AngularJS, mongo, express. It was hosted on heroku. The second iteration was based on meteor with everything real time. This is the version that's still running today.
The project has been running for many years. The farmer's contribution (14.90€/month) pays for the server, but with only 8 users the app is barely generating a tiny profit. I could, in theory, get a lot more users, the response to my cold calls was excellent.
But this project taught me something : being responsible for a professional tool that can crash at any moment is no fun.
The risk of the app crashing, the emails ending in spam, the reminder SMSs failing, etc .. is quite high, and the impact is that farmers can end up blocked in their work, and lose sales.
Today, i tend to prefer creating completely free side projects, because it stresses me less to have a program crash if said program is free to use. It's also forcing me to create very lean applications that tend to be frontend only. Therefore, they can be hosted for free indefinitely.
I'm still interested in the SAAS model, but I now keep in mind that being on call 24/7 limits you freedom. I live in Russia now, and would love to spend two weeks in the Trans-Siberian. But that's mean spending two weeks doing support from a train in the middle of nowhere. That's a no-go.
So, next time, I'll think a bit more before creating a product people will rely on.