Insights · 10 min read
A practical guide to custom software for small and medium businesses: when off-the-shelf stops working, what to build first, what it costs, and how to pick a vendor.
By GGP Editorial
Most small businesses do not need custom software until the day they do. That day usually arrives as a spreadsheet three people edit at once, a process only one person in the building understands, or a customer asking for something your off-the-shelf tools cannot do. When that happens the question stops being "should we use software" and becomes "should we buy something or build it."
I write this from the side of the people who build it. GlobeSoft has shipped custom systems for small and mid-size companies across several markets since 2018, from a financial management system to a property rental platform to a self-ordering and POS setup for restaurants. The notes below are the practical version of that work, aimed at a business owner or operations manager who has to make a decision, not at an engineer.
Packaged software works because it solves the common case. A small business becomes a custom software buyer when its case stops being common. You usually notice it one of four ways.
The workaround layer. You run an accounting tool, a CRM, a spreadsheet, and an email inbox, and your real process lives in the gaps between them. Someone copies data from one to the next by hand. Every manual step is a place where a number gets mistyped and a customer gets the wrong invoice.
The rules you cannot express. A rental business that charges late fees from a lease start date. A distributor whose prices depend on customer tier and order volume. A clinic whose schedule has to respect three kinds of appointment. Off-the-shelf tools either cannot model these rules or model them so awkwardly that staff quietly stop using the tool and go back to paper.
The customer-facing gap. Your customers want a portal to check a balance, book a slot, or see an order status. The packaged tool has no portal, or it has one with someone else's branding and someone else's data.
The integration wall. Two systems you already paid for will not talk to each other, and the vendor wants another subscription for a connector that half-works.
None of these are problems you fix by buying more software. They are problems you fix by building the small piece of software that is missing.
People hear "custom software" and picture an enterprise project with a hundred screens and a year-long timeline. For a small business it is usually the opposite. Most SME builds fall into one of these four shapes.
| Type | What it is | Typical trigger |
|---|---|---|
| Internal tool | A screen that replaces a spreadsheet or paper form | Errors, double entry, no audit trail |
| Customer portal | A login where customers see balances, orders, bookings | Customers emailing you for status |
| Integration | Code that connects two systems you already run | Manual copy-paste between tools |
| Workflow automation | Rules that move a job through steps on their own | A process that stalls waiting on a person |
The common thread is that the software encodes how your business actually runs instead of forcing your business to run the way a generic tool expects. That is the whole point. If your business works across markets, selling into more than one country, the same logic extends to currency and language, which a one-size tool handles poorly.
Buy when the packaged tool covers most of what you need and the missing piece is not where your margin comes from. Build when the missing piece is what your customers actually pay you for, or when the manual work is expensive enough to measure.
A rough way to size it: ten hours a week spent moving data between tools is around five hundred hours a year. Paying a person to do that costs real money every year, and the cost rises as you grow. A small system that removes most of that manual work pays for itself faster than most owners expect, which is why this is a business decision more than a technology decision. We go deeper on the decision in two places: build vs buy software and custom software vs SaaS.
Cost is the first question every owner asks, so here is the honest version. Custom software for a small business is priced by scope, not by a menu, but the scope for an SME is usually contained, which keeps the numbers smaller than the enterprise figures people fear.
A focused internal tool or a single integration often lands in the low to mid five figures. A customer portal with a login, a few screens, and payment or booking tends to be a mid five-figure project. Something with more moving parts, like a rental management system or a self-ordering and POS setup, is a larger five-figure to six-figure build depending on how much of the existing operation it touches.
Those are planning ranges, not quotes. The number moves with a small set of drivers.
| What drives the price | Why |
|---|---|
| Roles and screens | Each role needs permissions; each screen needs design and logic |
| Integrations | Connecting to accounting, payment, or email systems costs real engineering |
| Payments and compliance | Handling money or personal data adds security scope |
| How much process it touches | More workflow steps means more logic and more edge cases |
For the fuller picture of where the money goes, read our breakdown of custom software development cost.
One cost people forget is maintenance. After the build, budget somewhere around fifteen to twenty percent of the initial cost per year for hosting, monitoring, security patches, and the small changes every live system needs. It is not a hidden fee. It is the difference between software that runs for years and software that quietly rots.
The biggest risk in an SME project is not that the software fails. It is that you build more than you need and the project drags on. Three rules keep it small.
Write the one job the software must do in one sentence. "Let a tenant see their balance and pay rent." "Let staff enter an order and print a kitchen ticket." If you cannot say it in one sentence, the scope is not clear yet. Our guide on writing a software requirements document turns that sentence into a real spec.
List the steps your people take today, in order, including the ugly manual ones. The ugly steps are the ones the software should remove. Most of the value in an internal tool comes from deleting steps, not adding features.
Resist the second phase until the first one is live. Midway through a build, every owner wants to add "just one more report" or "a customer-facing view too." That is how a two-month project becomes a six-month one. Ship the small version, use it for a month, and the next set of priorities will come from real use instead of a guess.
For a small business the vendor choice matters more than the tech stack. You want a team that has built your kind of system before and can explain it in plain terms. Ask for a reference where you can actually speak to the person who uses the software, not just see a screenshot.
Check that they hand over the source code and that you own it. A small business that cannot walk away from its vendor is renting, not building. Ask how the handover and documentation work before you sign. We covered the interview side in what to ask a software company first, and the fuller process in how to hire a software development company. If you are weighing an offshore team against a local one, how to choose an offshore partner covers the parts people miss.
A realistic timeline for an SME build is shorter than most owners assume. A single internal tool or integration is often four to eight weeks. A customer portal is six to twelve weeks. The thing that stretches timelines is requirements changing mid-build, which is why the scoping step above pays for itself.
Three shapes this work actually takes, from projects GlobeSoft has delivered.
A financial management system for a business that had outgrown spreadsheets. The build pulled revenue, expenses, and reporting into one place so the numbers matched for everyone instead of existing in three slightly different versions.
A property rental system where tenants log in to see their balance and pay, and the landlord sees late payments and deposits without chasing anyone. The manual follow-up that used to eat the first week of every month went away.
A self-ordering and POS setup for restaurants, where orders from the table flow straight to the kitchen instead of being re-typed. Each of these is small in scope, and each removed a specific, recurring cost.
That is the point worth repeating. The useful custom software for a small business is usually a small system that removes a specific, recurring cost.
Do I need custom software, or can I use off-the-shelf tools?
If a packaged tool covers what you need and the manual workaround is tolerable, buy it. Build when the missing piece is where your revenue or your daily cost lives.
How much does custom software cost for a small business?
A focused internal tool or integration often lands in the low to mid five figures. A customer portal is a mid five-figure project, and a more involved system like a rental or POS platform runs larger. These are planning ranges; your number depends on scope.
How long does it take?
A single internal tool or integration is often four to eight weeks. A customer portal runs six to twelve weeks. Requirements that change mid-build are what stretch these numbers.
Who owns the software after it is built?
You should. Confirm in the contract that you own the source code and can take it elsewhere if you ever need to.
Do I need an in-house developer to maintain it?
No. Most small businesses keep a vendor on a small retainer for updates and fixes, which is cheaper than a full-time hire for one internal system.
Can a small business afford custom software?
Usually yes, because the build is small and it removes a recurring manual cost. The question is whether the recurring cost is bigger than the one-time build.
---
Small businesses rarely need a lot of software. They need one or two small systems that fit the way they actually run. If you are trying to decide whether that is worth building, get the scope and the number clear first, then decide.
Tell us what you are building and where you are today. We typically reply within 24 hours.
