Most of the software conversations we have with business owners start the same way. Someone has a spreadsheet that grew into a monster, or a handful of tools that don't talk to each other, and they want to know if it's time for something custom. Sometimes the answer is yes. More often the honest answer is "not yet," or "only for one piece of it."
Our founder, Chris Roufail, came to software from finance. He spent years in equity research and FP&A, then built finance processes from scratch at two startups and automated them. That background shapes how we look at this decision. It is a capital allocation question. You spend money now to save time and avoid mistakes later, and the job is to pick the option with the best return and the least risk over a realistic period.
When off-the-shelf wins
If your process looks like everyone else's, buy the software everyone else uses. Accounting, payroll, email, a basic CRM, online checkout and single-location scheduling are solved problems. The big products in those categories have had years and a lot of customers to work out the edge cases. You will not out-build them for a reasonable price, and you shouldn't try.
Off-the-shelf is also the better call when:
- You need it working next week, not next quarter.
- Your process is still changing. If you are still figuring out how you sell or deliver, locking that into custom software is an expensive way to find out.
- The subscription is small compared to the work it handles. A $50 a month tool that covers 90% of what you need is a good deal, even if the last 10% is annoying.
- Compliance is the product. Payroll tax filing and card processing belong with specialists.
The risk with off-the-shelf is gradual. You add a second tool to cover a gap, then a third, then part of someone's job becomes moving data between them. Each step makes sense on its own. Added up, they become a cost that never shows up as a single line on the P&L.
When custom wins
Custom software earns its price in two situations: when the way you work is part of why customers choose you, or when the work it replaces is large, repetitive and expensive. Some signs you are there:
- You have bent a product so far that it fights you. Custom fields on top of custom fields, workarounds written up in a shared doc, and one person who understands how it all fits together.
- People on your team spend a real share of every week copying data, reconciling reports or chasing approvals. In our experience, once that passes roughly 10 hours a week across the team, a build tends to pay for itself within a year or two.
- Your pricing, quoting or scheduling rules are specific to you, so they live in spreadsheets because no product handles them.
- You pay per seat for a large team to use a small slice of a big product.
The downsides are real. A build takes weeks, where a subscription takes minutes. Someone has to maintain it. And if you hire the wrong builder, you can end up with code nobody else can work on. Whoever you hire, ask who will support the system after launch, and how quickly they respond when something breaks.
The middle option to look at first
There is a third path that gets less attention than it should: keep the tools you have and connect them. Most business software has an API, which is a way for other software to read and write its data. A small custom layer can move records between your CRM, accounting system and project tool, build your weekly reports, and send the reminders that today depend on someone remembering.
This is usually where we would start. It costs less than a full build, it keeps the parts of your setup that already work, and it targets the specific places where time is lost. If the connections get complicated after a year or two, that is useful too. It tells you exactly what a full custom system would need to do. The examples page shows what a few of these look like in practice.
What each option costs over three years
Comparing a monthly subscription to a one-time build price is misleading. We use a three-year window because that is about how long a business process stays stable before it needs a real rethink.
Below is an illustrative comparison for a 15-person team. The figures are made up for the example and are not quotes. The point is which lines to include.
| Three-year cost | Off-the-shelf only | Connect existing tools | Full custom |
|---|---|---|---|
| Subscriptions | $21,600 | $21,600 | $0 to $5,000 |
| Setup or build | $3,000 | $15,000 to $25,000 | $40,000 to $80,000 |
| Hosting and upkeep | $0 | $9,000 to $18,000 | $15,000 to $30,000 |
| Staff time on workarounds | $67,500 | $13,500 | $5,000 |
| Rough total | About $92,000 | $59,000 to $78,000 | $60,000 to $120,000 |
The staff time line assumes 10 hours a week of manual work at a loaded cost of $45 an hour, cut to 2 hours a week by connecting tools. Change those assumptions for your business and the answer can flip. That is the reason to do the math rather than go on instinct.
Two things usually stand out when owners run this for themselves. First, staff time is the biggest number, and it never appears on an invoice. Second, full custom is only clearly cheaper when the manual work is heavy or the subscriptions are expensive. If neither is true, connecting tools is the better bet. You can see how we price each kind of work on the pricing page.
Questions to ask before you decide
- What does this process cost us today in hours and in errors? Put an actual number on it.
- Is our way of doing this an advantage customers notice, or just habit?
- Will this process look about the same in two years?
- Have we really tried configuring the tools we already pay for?
- Who will own and maintain a custom system once it's built?
- If the builder disappears, can someone else work on the code?
- What is the smallest version that would prove the idea is worth it?
If you can't answer the first question, start there. Track the hours for two weeks. It is the cheapest research you will ever do, and it often settles the decision by itself.
Common questions
Is custom software only for large companies?
No. Plenty of 5 to 50 person businesses have one process that is specific to them and eats hours every week. A focused build for that one process is often the right size, and the rest of the business can stay on standard tools.
How long does a custom build take?
On our projects, a focused custom system usually takes roughly 6 to 12 weeks from the end of discovery to launch. Connecting existing tools is often faster. A big replacement of a core system takes longer, and anyone promising otherwise is guessing.
Can I start with off-the-shelf and switch to custom later?
Yes, and it is often the smart order. Running a process on standard tools first shows you what you actually need. Just make sure you can export your data, because that is what the custom system will be built on.