Bespoke Software vs Off-the-Shelf: How to Decide

The decision is usually framed as cost. It is really about whether the process the software supports is something that makes your business different, or something every business does the same way.

The one-line test

Build when the process is your competitive advantage. Buy when it is not.

Payroll is not your advantage; buy it. Email is not; buy it. Accounting is not; buy it. The way you quote complex jobs, allocate work to a field team, onboard a client, or track compliance across sites might be - and if it is, off-the-shelf software will make you do it the way everyone else does, which is the same as giving the advantage away.

Most businesses have one or two processes that pass this test and a dozen that do not. The mistake in both directions is common: building an email system (never do this), or running the one thing that makes you money through a generic tool with workarounds.

What off-the-shelf gives you

• Working today. The value of this is underrated.
• A vendor who fixes bugs and adds features on their roadmap, not yours.
• Predictable monthly cost, at least until the seat count grows.
• A labour market that already knows the tool.

What bespoke gives you

• Screens that match the job, so training is minutes and adoption is not a project.
• No per-seat cost: the fortieth user is free.
• Integration with whatever you already run, on your terms.
• Ownership: the code, the data and the hosting are yours. The vendor cannot raise the price or retire the product.

And the costs: 6 to 16 weeks before it exists, a supplier relationship to manage, and an internal owner who will make decisions about it for years. We are transparent about this on our web application and CRM development pages, because a business that goes in expecting a purchase rather than a project is the one that ends up unhappy.

The five-year cost, honestly

For off-the-shelf: licence cost for the tiers you will actually need (not the ones you need today), times the seats you will have in year three, times five, plus the cost of the workarounds - the spreadsheet someone maintains, the hours re-keying data between systems.

For bespoke: the build (see the web application cost guide), plus hosting at £30 to £150 a month, plus maintenance at 10 to 15 percent of the build per year, plus a budget for change.

Run both numbers. In our experience, bespoke wins on cost alone somewhere between year two and year four for businesses with twenty or more users, and never wins on cost for five users. Below twenty seats, build only if the process argument is strong.

The trap in "bespoke" quotes

Some "custom" software is bespoke in name only. The tells:

• Mandatory hosting on the agency's servers. If you cannot take the system elsewhere, you do not own it.
• A licence fee. Custom software does not have a licence. If it does, you have bought a product with a custom skin.
• Support retainer as the only route to fixes. Bugs in the first months are the agency's responsibility; a retainer is for new work.
• No repository access. The code should sit in a repository you control from day one, not be "handed over" in a zip file at the end.

Ask each of these before you sign. The answers separate agencies that build software from those that rent it.

A worked decision

A training provider runs courses across several venues, with trainers, materials, bookings, certificates and compliance records. Their accounting is standard: buy. Their email marketing is standard: buy. Their course scheduling, trainer allocation and certificate issuing is the business, and every off-the-shelf tool they tried made them do it the tool's way: build that, integrate it with the accounting and email systems they already have, and stop maintaining the four spreadsheets that held it together.

The hidden cost of off-the-shelf: workarounds

Licence fees are visible. The cost that is not is the work done around the software because it does not fit: the spreadsheet that holds what the system cannot, the re-keying between two tools that do not talk, the weekly report assembled by hand from three exports, the process step that happens in email because the system has no place for it. Add up the hours. In businesses of twenty to fifty people, we regularly find one full-time equivalent absorbed by workarounds, which at UK salaries is £25,000 to £40,000 a year - more than most builds cost.

The hidden cost of bespoke: attention

A built system needs someone to care about it. Not full time, but consistently: deciding what changes are worth making, testing them, keeping the supplier relationship healthy, watching for the day a dependency needs upgrading. If nobody in the business will do this, the system will drift out of date and the vendor who told you to buy will have been right. Name the owner before you sign.

A decision matrix

• Standard process, few users: buy. Always.
• Standard process, many users: buy, but check the seat maths at year three.
• Unusual process, few users: buy the closest fit and tolerate the gaps, unless the gaps are where the money is.
• Unusual process, many users: build the core, buy the commodity around it, integrate.
• Regulated data with specific handling rules: lean towards building the part that handles it, so the rules are yours and testable.

Timing

The right moment to build is usually after you have outgrown a bought tool, not before you have used one. A year on off-the-shelf software teaches you exactly what your process needs, which makes the brief for a custom build far better - see how to brief a software agency. Building first, from a blank page, produces systems that model how the business imagines it works rather than how it does.

Questions to ask before deciding

• If we changed this process tomorrow, would customers notice? If yes, it is probably core.
• How many hours a week go into working around the current tool? Multiply by fifty and by a loaded hourly cost.
• What will the seat count be in three years, and what does the vendor's next tier cost?
• Who would own a built system, by name, and do they want to?
• Which off-the-shelf tools have we actually trialled, and what specifically did not fit?

Written answers to those five make the decision for most businesses, and they make a far better opening to a conversation with an agency than "we think we need a system".

That pattern - buy the commodity, build the core, connect them - is the answer for most businesses we talk to. If you want a second opinion on where the line falls in yours, send us a description of the process and we will tell you which side of it we think you are on.