Off-the-shelf or custom software: how to decide in one hour

Should you buy off-the-shelf software or build your own business tool? The question of off-the-shelf or custom software comes up in almost every first meeting we have, and it is badly settled by opinion. The answer comes down to 3 questions: is your process standard, does your competitive edge depend on it, and what does each option really cost over 3 years. Here is a method to decide in one hour, decision table included, with the orders of magnitude we practise.

The 3 questions that decide it

First question: is your process standard? Payroll, accounting, appointment booking, email look alike from one company to the next: the market covers them very well, at a price nobody can beat by building. If your way of working fits inside the configuration options of an existing product, buy it.

Second question: does your competitive edge run through this process? If the way you qualify a file, price a quote, run a project or serve a client is precisely what makes you win, pouring it into the mould of generic software erases what sets you apart. That is where custom is justified, and only there.

Third question: what does each option really cost over 3 years? Not the sticker price, the full cost. That is the object of the next section, and it is almost always the worst-asked question.

The real cost over 3 years, subscriptions included

On the off-the-shelf side, the honest calculation adds five items.

On the custom side, the calculation is shorter: build cost, plus running costs, hosting, maintenance, evolutions. With us, a business tool starts at €6,000 excl. VAT as a Sprint for a first scope usable in production, and the Care maintenance plan starts at €250 excl. VAT per month. Our public calculator runs the full comparison in 5 minutes with your own numbers, no email required.

Draw both curves: there is almost always a crossing point. Before it, the market wins; after it, every month of per-seat subscription pays for a tool you still do not own. The decision becomes a chart to read, not an argument to win.

What tips the balance towards the market

• The process is regulated and standardised: payroll, accounting, filings. The market follows legal changes for you, which a custom tool will not do without budget.
• Volume is low and stable. A tool used by two people once a week does not justify a build.
• Nobody is available to decide. A custom project needs one person who arbitrates; without them, it stalls.
• You do not yet know how you want to work. Off-the-shelf software imposes a way of working; sometimes that is exactly what a young team needs to structure itself.

What tips the balance towards custom

• The process is your signature. Clients choose you for it, and it resembles no other.
• You stack three tools and two spreadsheets for a single chain of work. Every boundary between tools is a re-entry, a possible error and lost information.
• The number of users is growing. Per-seat pricing becomes a tax on your growth.
• Your data must stay with you, with precise access rules, for confidentiality or professional-duty reasons.
• You want to plug AI agents into your rules, not into a publisher's average rules.

The one-hour decision table

The method fits in one meeting. List your five most important processes, not fifty, five. For each, three columns.

The reading is immediate: standard, low-friction processes stay on the market; differentiating processes riddled with re-entries are the candidates for custom; the rest waits. One hour is enough to escape the trick question "should we rebuild everything?" and leave with a roadmap per process, in priority order.

The third way: off-the-shelf at the core, custom at the edges

In practice, the right answer is rarely all one or all the other. The pattern that works: keep the market for standard functions (accounting, payroll, office tools, email), build custom the layer that sets you apart (the cockpit, the client portal, the quoting engine, case tracking), and connect the two so data flows without re-entry.

That is exactly the architecture of Stratedge OS: your current tools stay in place, the platform connects them, cleans up the data and runs the custom blocks on top. We do not throw away what works; we build what is missing.

The mistakes we see most often

1. Rebuilding off-the-shelf software. Rebuilding an invoicing tool or an email client is a guaranteed loss. Custom should only carry what the market cannot do for you.
2. The forty-page specification. It raises the price for functions nobody will open. A first scope should only cover what will be used every week.
3. Ignoring data migration. Years of clients, quotes and documents to migrate are a project of their own, to be priced from the start.
4. Choosing by the demo. A demonstration shows the best case. Test the market tool on your worst case: the complicated file, the client with three addresses, the deposit invoice.

Where to start

If three processes in your table land in the custom column, do not launch them together. Start with the one that crosses the most re-entries: its gain will be visible fastest. An Express Diagnostic at €500 excl. VAT, deducted from what follows, delivers within a week a written, costed scope and a clear recommendation: market, custom or hybrid. You decide with numbers, with no commitment on what comes next.

10 questions a serious quote answers

If the table sends a process towards custom, the next step is a quote. Judge it less on its amount than on what it writes down. A serious quote answers these 10 questions, in black and white.

1. The exact scope of the first block, and above all what is excluded.
2. The milestones and what each one delivers. A milestone is a usable deliverable, not a progress meeting.
3. The integrations planned, which tools are connected, in which direction data flows, who provides access.
4. Data migration, what is migrated, in what state, who cleans it.
5. Ownership. Code, data, hosting accounts and domain names in your name.
6. Hosting, where, with whom, with which backups, at what monthly cost.
7. Training and handover, how many hours, for whom, with which written material.
8. Warranty, for how long defects are fixed without invoicing.
9. Maintenance, what it covers, what it excludes, the commitment period and the notice.
10. Payment terms, the percentages, the dates, and what triggers each payment.

A provider who answers all ten on one page is selling you a system. A provider who answers "we will see along the way" is selling you hours.