One order, three tabs, and a lot of retyping
A cabinet shop takes an order through its website form on a Thursday afternoon. The office manager copies it into the scheduling tool, types it again into the invoice, and pastes the address into the crew's route sheet. That is twenty minutes of retyping, three chances for a typo, and it happens several times a day.
Nothing here is broken, which is exactly why it never gets fixed. The retyping is not the only cost: every manual hop is a place where a detail can drift or a job can stall. Sooner or later every owner asks the same question about work like this: buy software for it, or have something built.
The real question is not build or buy
The more useful question is whether the problem you are solving is common to everyone or particular to you. Accounting, payroll, email and card processing are the same for every business, and off-the-shelf tools serve them well.
How you quote, schedule crews, follow up and collect is particular to your shop. That is where bought tools start to pinch, and where custom work starts to pay.
There is also a third option that the build-or-buy framing hides: keep every tool you have and build the connections between them. Much of what gets sold as new software is really a missing handoff between two tools you already pay for.
When buying wins
Buy when the problem is a commodity and maintaining the tool is somebody else's full-time job. The signs that you are in commodity territory are easy to read.
- The problem is identical across thousands of businesses: payroll, accounting, document signing, card processing.
- A standard tool covers your actual workflow without forcing awkward workarounds.
- The monthly price stays reasonable as seats and usage grow.
- You can leave: your data exports in a format another tool can read.
Rent the commodity and give it no more thought than it earns. Your edge was never going to come from rebuilding accounting software, and the maintenance burden alone would erase any savings. The subscription is not a failure of imagination, it is the market doing your maintenance for you.
When building wins
Build when the workflow is particular to you and every bought option forces you to work its way. These are the signs we see most often in small operations.
- The business runs on a spreadsheet only one person fully understands.
- Your team re-enters the same information into two or three systems every day.
- You pay for overlapping subscriptions and use a small slice of each.
- A tool exists, but it forces a process your customers can feel.
Custom does not have to mean a big company project with a big company budget. Our builds typically run $1,000 to $5,000, take a few days to a month, and run on boring, reliable web infrastructure. The price arrives as a written quote before anything starts, so it is a decision, not a discovery.
The third option: connect what you already run
Before you replace anything or build anything, look at the handoffs between the tools you already run. A web form can create the job record by itself, and the invoice can go out the moment a job closes. The review request can follow by text the next morning.
Glue work keeps the tools your team already knows and removes the retyping between them. It is usually the cheapest option on the list and the least disruptive. Exceptions route to a person with context, every automated action is logged, and nothing that works today gets replaced.
Start the audit at the moments money changes direction: a lead arrives, a job closes, an invoice goes out, a payment lands. Each of those moments either triggers the next step by itself or waits for a person to remember it.
Count the full cost, not the sticker price
Bought software looks cheap because the price arrives monthly, and custom looks expensive because it arrives once. Put both on the same three-year line before you decide anything.
Suppose the niche subscription that fits your workflow costs $150 a month for the seats you need. That is $1,800 a year and $5,400 over three years, before you count the hours spent working around its gaps. A $3,000 custom build sits under that line, and you own the result at the end.
Hours are the half of the cost nobody writes down. If retyping takes forty minutes a day at $20 an hour, that is about $13 a day and roughly $3,300 across a working year. Whichever option removes those minutes is quietly worth that much, every year it keeps working.
None of this says custom always wins, it says the sticker price settles nothing on its own.
A simple decision test
Run each candidate problem through five questions before you spend a dollar on it.
- Is the problem common to every business or particular to yours? Common leans buy, particular leans build.
- Does a bought tool cover the workflow without changing how customers experience you? If yes, buy it.
- Is the real problem a missing handoff between tools you already like? Connect them instead of replacing them.
- Does the three-year arithmetic, hours included, clearly favor one side? Trust the math over the demo.
- Can you leave later, with your data, documentation and logins intact? If not, keep looking.
Ownership is the tiebreak
Whichever way you go, insist on ownership: with bought tools that means exportable data and terms you can walk away from. Ask the exit question while everyone is still friendly, because the answer only gets worse later.
With custom work it means the finished build, the documentation and a proper handover belong to you. That is how we structure every project: a written quote and statement of work before anything starts, with three rounds of revisions included. Hosted builds carry a 99.9% uptime target, and monthly plans cover updates, maintenance and uptime after launch.
If you are not sure where your list lands, our workflow walkthrough is free and fixed in scope. You get a ranked plan with an impact estimate, and when buying is the right answer, the plan says buy.