Custom software for a small business typically costs somewhere in the $2,000–$10,000 range for a focused automation — a single process replaced or streamlined — and $5,000–$50,000+ for a larger internal tool or customer-facing system. Those ranges are wide because "custom software" covers everything from a form that emails you when a customer submits an enquiry to a platform that manages your entire operation.
The most important thing to understand up front: the price isn't mainly set by the technology. It's set by how well-defined the problem is, how many things the software needs to talk to, and whether the scope stays fixed through the build. Two businesses asking for "the same thing" can end up with quotes that are several times apart for reasons that have nothing to do with complexity.
What You're Actually Paying For
When you commission custom software, the build itself is roughly what you'd expect: someone writes code that does the thing. But the visible part of the work — the screens, the buttons, the reports — is often a smaller share than clients expect.
The bigger costs are the unglamorous remainder:
Understanding the problem. A developer who ships the wrong thing efficiently is expensive. Most of the early work — discovery calls, mapping the current process, writing the spec — looks like talking and document editing. It's not. It's the difference between software that fits and software you abandon six months later.
Edge cases and testing. Every real process has exceptions. The customer who places an order with a weird character in their name. The report that breaks on the last day of the month. The data that was entered inconsistently before the system existed. Finding and handling those cases takes time, and the cases that take the longest to fix are always the ones nobody thought to mention during discovery.
Handover. Documentation, training, making sure whoever runs the tool day-to-day can actually use it without calling the developer. Often undervalued, always regretted when skipped.
The Three Price Tiers
Focused Automation: $2,000–$10,000
A single process, automated. A web form that captures job requests and emails the right person. A script that pulls data from your booking tool and writes it to a spreadsheet, eliminating the manual copy. A simple dashboard that shows today's outstanding invoices.
These are contained, well-defined, and fast to build when scoped properly. They're also the best place to start: lower risk, faster payback, and they teach you whether this approach actually works before you commit to something larger.
Internal Tool: $5,000–$50,000+
A system your team uses to run operations — managing jobs, tracking client history, generating quotes, scheduling work. This is where most small business custom software sits.
The range is wide because the number of features, user accounts, integrations, and the state of your existing data all move the number significantly. A job management tool for a two-person team with clean data is a different project from one for a twelve-person team that needs to pull in records from three existing systems.
Customer-Facing System: typically at the higher end
Software your customers use directly — a client portal, an online booking flow with custom logic, a self-service tool that connects to your backend. This tier is more expensive because it needs to handle users you don't control: browsers you didn't test on, input you didn't anticipate, security requirements that don't apply to internal tools.
A Shopify app I built is used by over a hundred merchants and has processed a meaningful volume of transactions. The customer-facing layer — authentication, error handling, the edge cases that turn up when real users arrive — took as long as the core functionality.
What Makes the Same Project Cost Several Times More
Integrations. Every system your software needs to talk to — your accounting tool, your CRM, your email platform, your booking system — is a dependency. Each one needs to be understood, connected, and tested. Each one can change without warning, breaking the connection and requiring a fix. A single integration can move a quote noticeably; several can dominate it.
User accounts and permissions. "People need to log in" sounds simple. It isn't. Authentication, session management, password resets, roles, what each role can and can't see — this is a substantial block of work that most clients underestimate.
Migrating messy data. If you're replacing an existing system and need to bring the old data across, the quality of that data matters enormously. Clean, consistent records import quickly. Years of spreadsheets with inconsistent formats, missing fields, and duplicate entries require cleanup before migration — and that cleanup takes time. Every business thinks their data is cleaner than it is. Plan for the cleanup to take longer than expected.
Scope growth mid-build. "While we're at it, could we also..." is the most expensive sentence in software. A feature added mid-build costs more than the same feature scoped at the start, because the developer has to understand an existing system, integrate the new thing cleanly, and retest things that were already done. Write the full wish-list before work starts. Agree what's in scope and what's phase two. Then hold the line.
How to Keep the Cost Sane
Start with the single most painful process. Not the complete vision. Not everything you'd love the software to do eventually. The one thing that causes the most actual pain each week. Build that, get it working, and let it prove its value before expanding.
Insist on a written, fixed scope before work starts. If a developer won't commit to a scope document, keep looking. A fixed scope means you know exactly what you're getting and what it costs. Hourly billing on a vague brief is how projects run over budget and leave both sides frustrated.
Hold phase two until phase one has earned it. It's tempting to add more scope the moment the first version is working. Resist it. Let the first phase settle. See what actually gets used and what doesn't. The things you thought you needed in phase two often look different after six months of real use.
When Custom Is the Wrong Answer
I build custom software, so admitting this costs me potential work — but here it is: custom is often the wrong answer.
If there's an off-the-shelf tool that covers your needs at $50–200/month, use it. Xero, HubSpot, Airtable, Calendly, and dozens of others exist because the problems they solve are common. Custom software makes sense when the process is genuinely unusual, when the off-the-shelf options don't connect to your other tools, or when the cost of workarounds has exceeded what a proper solution would cost. That bar is higher than most people assume.
Sometimes the right answer is to simplify the process first. A process that's complicated because it evolved over years without anyone questioning it isn't a good candidate for automation — automating a bad process makes a fast, reliable bad process. Simplify first, then consider whether software helps.
For a structured comparison of when to buy versus when to build, read Off-the-Shelf vs Custom Software. If you're at the point of wondering whether you've even outgrown your current tools, 7 Signs Your Business Has Outgrown Spreadsheets is a useful starting point.
If you've already decided custom is the right direction and want to talk through what a fixed-scope project would look like, here's how I work.
