Most Custom vs Off-the-Shelf Software Guides Get the Question Wrong
Open any article on custom vs off-the-shelf software and you’ll find the same tidy table: bespoke is flexible but expensive, SaaS is cheap but rigid. All true. None of it helps you decide what to do on Monday morning.
The pros-and-cons format assumes you already know which side of the line you’re on. Most UK SMEs don’t. They’re staring at a £40k development quote on one screen and a £180-per-seat-per-month SaaS subscription on the other, and the honest answer to “which is better?” is: neither, until you’ve answered a different question. Which is right for this workflow, this team, this moment?
By the end of this post you’ll have a scoring rubric you can apply to your own situation in about twenty minutes — and a clearer view of when the hybrid middle ground is actually the right call.
Four Scenarios Make Custom Software Worth the Cost
Custom software is expensive to build, expensive to maintain, and expensive to replace. It earns that cost in four specific situations.
1. Your workflow is a genuine differentiator. If the way you process orders, route engineers, or price quotes is part of why clients choose you over competitors, no SaaS vendor has modelled it — because if they had, it wouldn’t be a differentiator any more. Forcing a unique process into a generic tool usually means flattening exactly the thing that makes you valuable.
2. The software is the product. If what you sell or deliver to clients is itself a digital system — a portal, a calculation engine, a data service — then the software isn’t an internal tool, it’s the asset. Building on someone else’s platform means renting your own product back from them.
3. Sector-specific regulation. FCA-regulated firms, CQC-registered care providers, SRA-regulated solicitors and similar all face data handling, audit trail and reporting requirements that off-the-shelf products tend to address generically or not at all. Where you need defensible evidence that a control existed on a specific date, bespoke usually wins.
4. Scale economics have flipped. Per-seat SaaS pricing is brilliant at 15 users and brutal at 150. Run the five-year total cost — licences, integration fees, admin overhead — against a one-off build plus maintenance. At a certain user volume the curves cross, and ignoring that crossover is how SMEs end up with seven-figure annual SaaS bills.
Four Scenarios Make Off-the-Shelf Software the Right Default
The reverse is just as clear-cut. Building software you didn’t need to build is one of the most expensive mistakes an SME can make.
1. The workflow is a commodity. Payroll, expense management, basic CRM, email marketing — thousands of businesses do these the same way, and mature SaaS products have spent a decade encoding best practice into their default settings. Building your own version is rebuilding a wheel that already rolls.
2. The requirement isn’t mature. If you can’t describe in detail what the software should do, building it now locks in today’s half-formed assumptions as tomorrow’s expensive constraints. Off-the-shelf lets you learn what you actually need by using something — then build later if the gap persists.
3. You’re a small team. Custom code needs someone to own it. Updates, security patches, bug fixes, the developer who built it leaving — all of that is manageable at 100 staff and disproportionate at 12. SaaS vendors absorb that cost across thousands of customers.
4. The clock is the constraint. A live SaaS tool configured in a fortnight beats a six-month build when a client onboarding, a funding round or a cash flow squeeze is the actual deadline.
Most UK SMEs Need Off-the-Shelf Platforms With a Custom Integration Layer
Most UK SMEs we work with sit in neither camp. They don’t need a full bespoke system, but they’re losing hours every week to a SaaS stack that doesn’t talk to itself.
The pattern that works: pick the best-in-class SaaS product for each domain — Xero or similar for finance, HubSpot or Pipedrive for CRM, a sector-appropriate ops platform — and wire them together with a lightweight custom integration layer. Quotes flow into invoices without re-keying. New clients appear in the ops system the moment the deal closes. Management reporting pulls from one place because the data has been reconciled, not because it all lives in one monolith.
This preserves what makes SaaS valuable: vendor roadmaps, third-party support, lower per-module cost, the ability to swap out a component when something better arrives. What it removes is the manual labour and the silos.
This is the territory Agile Operations is built for — designing and maintaining the integration layer so the business owns the logic and the data flows, without owning the whole stack.
The honest caveat: the integration layer itself needs governance. Undocumented webhooks and a Zapier account shared with three ex-employees can become their own form of technical debt fast. Treat the integration layer as a product in its own right — documented, monitored, owned — or it’ll quietly become the fragile bit you’re afraid to touch.
A Build-vs-Buy Scoring Rubric Surfaces the Dominant Signal
Score each criterion 1, 2 or 3 against the guide below, multiply by the weight, and total the column.
| Criterion | Weight | Scoring guide |
|---|---|---|
| Process uniqueness | ×3 | 1 = standard practice; 2 = some unusual steps; 3 = core differentiator |
| Regulatory specificity | ×3 | 1 = general GDPR only; 2 = some sector rules; 3 = FCA/CQC/SRA-level audit needs |
| Integration complexity | ×2 | 1 = standalone use; 2 = 2–3 systems to join; 3 = 4+ systems, real-time |
| Team size & dev capacity | ×2 | 1 = under 20 staff, no dev; 2 = 20–100, some tech capacity; 3 = 100+ with internal dev |
| Time-to-value pressure | ×2 | 1 = need it live in weeks; 2 = a few months acceptable; 3 = 6+ months fine |
| 5-year total cost comparison | ×2 | 1 = SaaS cheaper; 2 = roughly even; 3 = custom clearly cheaper at scale |
| Requirement maturity | ×1 | 1 = still exploring; 2 = mostly defined; 3 = fully specified and stable |
Thresholds: – 0–20: Off-the-shelf. Buy, don’t build. Spend the saved budget on configuration and training. – 21–32: Hybrid integration layer. You’ll get more from joining tools together than from replacing them. – 33–45: Custom build warranted. The numbers say the bespoke route earns its cost — go in with eyes open.
The rubric surfaces the dominant signal, not the final answer. If you land squarely in the hybrid band, a conversation with an integrator is still the right next step — the score tells you what kind of conversation to have.
If you’d like a second opinion on a specific build-vs-buy decision, that’s the kind of conversation Agile Operations is built for. No pitch, just a structured look at your situation.