“I could probably build something that does what we need.”
If you’ve spent enough time fighting with spreadsheets or entering the same information in three places, you’ve probably thought about it. Maybe you know somebody who writes software. Or you’ve been playing around with AI and already have a customer screen that looks pretty good.
The temptation makes sense. Build it your way, leave out what you don’t need, and stop paying somebody else every month.
But there’s a lot of work hiding behind that order screen.
At PriceIt Software, we’ve given at least 20 live demos over the years to companies that had already built their own shop management software.
They’d put time and money into it. They knew their shops, knew what they wanted, and had gotten a system up and running. But by the time they were talking to us, they’d decided it wasn’t worth continuing to maintain it.
The reasons were familiar: fixing problems, keeping up with changes, supporting employees, and depending on one person who knew how everything worked. What started as a way to make running the shop easier had become another responsibility they wanted off their plate.
We’ve spent decades developing software for decorated apparel shops, and we understand how that happens. Much of the work involves things you’d never think about until a customer needs them—or something breaks.
Before you start building your own, here are ten things to think through.
1. Your shop is more complicated than you think
Customers, quotes, orders, invoices. That sounds manageable.
Then a customer orders 144 shirts, supplies 24 of them, wants embroidery on some and screen printing on the rest, and needs two separate delivery dates. They’ve paid a deposit, but now they’re changing the quantities.
Your employees already know how to deal with this stuff. Getting software to handle it properly takes work.
And that’s one order. Next week, somebody will bring you a different problem.
You don’t have to build every possible feature on day one. But the things your shop does every day need to work without somebody keeping a separate notebook to explain what’s actually happening.
2. Getting it to work once doesn’t mean it’s ready
You enter an order, print an invoice, and everything looks right. Great.
Now change the quantity after taking a deposit. Delete a line. Add a discount. Have two people open the order at the same time.
Still working?
How about when somebody loses their connection halfway through saving?
This is where a lot of the development time goes. You have to check calculations, fix problems, and make sure fixing one thing didn’t break something else.
If you skip that work, your employees will find the problems for you. Usually when they’re busy.
3. AI doesn’t relieve you of knowing what you’re doing
AI can help write code. It can get you past a problem that would have taken hours to figure out on your own. That’s useful.
It can also give you something that appears to work, and you may not know enough about the code to spot what’s wrong.
An incorrect total is easy to catch when it’s obviously ridiculous. A small pricing error that only happens under certain conditions could sit there for months. By then, you’ve quoted jobs and sent invoices based on it.
You still need somebody who understands the system well enough to check the work and fix it. “The AI wrote it” won’t help much when a customer wants to know why their invoice is wrong.
4. You’ll need somewhere to run it
Maybe you’ll put it on a computer at the shop. Maybe you’ll rent a server.
Either way, there are things to take care of.
Computers fail. Drives fill up. Internet connections go down. Updates need to be installed. Employees need access, and former employees need their access removed.
You’ll also need to think about security: who can see customer information, who can change prices, and how people connect from outside the shop.
Renting a server doesn’t necessarily mean somebody else handles all of that. It depends on what you’re paying them to manage.
If the system is down when everyone comes in Monday morning, who gets the call? And do they know how to fix it?
If the answer is you, leave some room in your Monday schedule.
5. You need to know whether your backups actually work
“We have backups” is a good start.
Have you ever restored one?
You need to know where they’re stored, how often they run, and how far back you can go. Keeping a copy on the same machine won’t do much good if that machine is lost or damaged.
There’s also a difference between recovering from a failed server and recovering information somebody accidentally deleted three weeks ago.
These are things to work out while everything is running. During an outage, you’ll have enough to do without discovering that your backup isn’t what you thought it was.
And getting the data back is only part of it. You need to know how long the shop could be without the system and what everyone will do in the meantime.
6. Connecting to other services takes ongoing work
At some point, you’ll probably want to send invoices to accounting, pull in supplier information, or bring orders over from an online store.
That can save a lot of time once it’s working.
Getting there means figuring out how the other system expects to receive information, what to do when something is missing, and how to avoid sending the same thing twice.
Then the other company changes something, and you’re back working on it.
You don’t control their updates or their outages. But when the connection stops working, your staff still needs an answer. They don’t particularly care whose end the problem is on. They need to get the order processed.
7. Your employees will need help
Somebody can’t open an order. Somebody entered a payment incorrectly. Somebody got a new computer, and now something won’t print.
You’re going to hear about it.
Some calls will be bugs. Others will be training questions or problems with a computer. You’ll have to stop long enough to figure out which.
It’s easy to overlook this when you’re planning the project. You’re thinking about what the software will do, not how many times you’ll be called away from your own work to help someone use it.
But once people depend on it to do their jobs, supporting them becomes part of yours. So does training the next employee who comes through the door.
8. A lot of the work happens when you should be home
During the day, you’re running a shop. There are customers to deal with and orders to get out.
So you work on the software after dinner.
Or you come in Saturday because it’s easier to make changes when nobody else is using it. You figure it’ll take an hour or two. Then something goes wrong, and there goes Saturday.
Occasionally doing that may not bother you. Doing it regularly gets old, especially for the people waiting for you at home.
And taking a vacation gets harder when you’re the only person who can fix the system.
You may enjoy building it. Just be honest about where those hours are going to come from.
9. What happens if the person who built it leaves?
Maybe your employee is good with computers. Maybe your friend has offered to build it for a reasonable price.
Before the whole shop depends on their work, talk about what happens if they’re no longer available.
Do you have the source code? The passwords? Access to the server account? Is anything documented well enough for somebody else to take over?
Hiring another programmer doesn’t guarantee a quick handoff. They may need considerable time just to understand what the first person built. You’ll be paying for that time before they make a single improvement.
That applies if you’re the developer, too. You should be able to take a week off without being on call for every software problem.
10. Your time belongs in the price
If you’re doing the work yourself, it’s tempting to count only the money coming out of the bank account.
A server bill here. A development tool there. Looks pretty cheap.
But say you spend five hours a week maintaining the system and helping people with it. Over a year, that’s 260 hours—more than six full workweeks.
You might decide that’s time well spent. But count it before you compare the cost of building with the cost of buying.
Include the time you spent learning, testing, fixing mistakes, and adding the things you didn’t realize you needed. Otherwise, you’re comparing a subscription price with a project you’ve assigned free labor.
It’s the same principle behind job costing and profitability: you need to account for the work going into something before you can decide whether it’s making financial sense.
What does it really cost you?
Put the subscription price next to an actual job.
Depending on your pricing, costs, and the software plan you choose, the profit on one 24-piece order can pretty much cover a month’s subscription to shop management software. Take a look at monthly software pricing and compare it with what your shop actually earns on that order.
How many hours would you spend building, fixing, and supporting your own system to save that amount?
Those hours have to come from somewhere. When you already spend your working day running a shop, they often come from the people at home.
Your spouse spends another evening alone while you finish “one more thing.” Your children want you to come outside, but you’re trying to figure out why an update broke the order screen. You planned to spend Saturday with your grandchildren, and now you’re at the shop getting the system running again.
Occasionally, that happens to any business owner. The question is how much more of it you want to take on.
Even when you make it home, an unresolved problem can follow you. You’re checking your phone over dinner or thinking through a fix while somebody is trying to tell you about their day. Your family gets your time, but only part of your attention.
It’s easy to tell yourself you’ll catch up with them when the software is finished. Meanwhile, there’s another problem to fix or another feature the shop needs. The kids keep growing, and the afternoon your grandchild wanted to spend with you has passed.
There are shops where custom software makes sense. They have specific needs, a budget for development, and somebody whose job includes keeping it running. Building a small tool to solve one annoying problem can make sense, too.
But taking responsibility for the system your entire shop depends on is a much bigger commitment. Those companies that came to us for demos had already found that out. They’d built their software. They were looking for a way to stop spending their time on it.
A subscription comes with a bill you can see. The cost of your own time is easier to overlook, but your family feels it.
If the profit from a 24-piece order can cover the month, think carefully about how many evenings and weekends you’re willing to give up to save it.
