A finance director at a Sheffield-based manufacturing business sent me a Slack message a few weeks ago. They were stuck. Their warehouse operations had outgrown their off-the-shelf inventory system and they were weighing whether to bolt on customisations, switch platforms, or finally build the custom system their ops team had been asking for since 2023. The quotes they’d received ranged from £40,000 to £480,000. The variance alone told them something was wrong with how they were thinking about it.

We spent a couple of evenings on it and worked through a proper framework. What’s below is basically that conversation written down.

The Question Has Changed

Build vs buy used to be simple. SaaS won most of the time because building anything custom was painfully slow and expensive. Then cloud got cheaper, AI tooling cut development time by 30 to 40 percent on routine work, and the calculation shifted.

Off-the-shelf is still fine for most things. CRM, accounting, HR, basic project management. Buy these. Don’t even think about custom.

But for anything that touches your competitive moat (how you fulfil orders, price products, route logistics, onboard customers), the build case is stronger in 2026 than it has been since the early 2010s. The talent has improved. The tooling has matured. The cost-to-build for a focused custom platform has dropped meaningfully.

So the real question now is more nuanced. Where exactly does custom pay back, and where is it still a money pit?

When Buy Is Almost Always Right

A few honest signals you should buy, not build.

If thousands of other businesses solve the exact same problem the same way, there’s a SaaS for it. Don’t reinvent the wheel. Payroll, time tracking, expense management, video calls, file storage, generic email marketing. Market is saturated. Anyone building custom here is wasting money.

If your team won’t use it consistently anyway, you don’t need a custom build. You need adoption training. Buying a tool people will use beats building a perfect tool that sits unused.

If the problem changes monthly with regulations or market shifts, SaaS handles updates so you don’t have to. Tax software is the classic example. The vendor absorbs the cost of staying current with HMRC and FCA changes.

If you’re pre-revenue or pre-Series-A, build only if it’s your actual product. Internal tools should stay off-the-shelf until you have the cash and team to support custom work properly.

When Build Starts Making Sense

The signals to consider going custom are different.

If your process is genuinely different from how other businesses operate, you’ll end up shoehorning yourself into a SaaS template that costs you efficiency every day. That hidden cost adds up.

If the SaaS subscription cost is creeping past £80,000 a year for a tool used by a focused team, custom starts becoming financially comparable. A £200,000 build amortised over five years is £40,000 a year of capitalised cost. Plus you own the IP.

If your competitive advantage runs on data flows that SaaS vendors won’t customise for you, build. The classic example is logistics. The pricing logic, the route optimisation, the inventory rules. These are your actual edge. Renting them from a vendor weakens you.

If integrations across multiple SaaS tools have become their own nightmare, sometimes a single custom platform replaces five subscriptions and a Zapier mess. The ROI here is often hidden but real.

A lot of UK software development companies can build this kind of focused custom platform in 6 to 9 months for between £150,000 and £400,000, depending on complexity and partner choice. That’s a meaningful range, so vendor selection matters.

The Hybrid Reality Most UK Buyers Land On

In practice, most UK businesses I work with end up somewhere in the middle. They keep SaaS for the commodity stuff and build custom for the 20 percent of operations where they have a real edge.

A typical 2026 stack for a mid-sized UK company looks like this. Xero or QuickBooks for accounting. HiBob for HR. Slack for comms. Off-the-shelf CRM, usually HubSpot or a Salesforce-lite setup. Then one custom platform that runs whatever makes them money, whether that’s bespoke order processing, a customer portal, a manufacturing dashboard, or a logistics optimisation engine.

That hybrid model is honestly the right answer for 80 percent of UK growth-stage businesses. Run costs stay reasonable while you build a defensible operational moat where it counts.

A Costs Reality Check

Custom software pricing in the UK varies more than people realise. Worth knowing the ballpark before you take a quote at face value.

A focused custom platform built by a regional UK agency or a hybrid UK-offshore firm typically costs £120,000 to £300,000 for the initial 6-9 month build. Ongoing maintenance runs 15 to 20 percent of build cost annually.

A London-based consultancy will quote 2 to 3 times those figures for the same scope. Sometimes the quality difference justifies it. Often it doesn’t.

Pure offshore quotes can land at £60,000 to £150,000 for the same scope, but you take on more management overhead and the quality range is wider.

Worth comparing a few of the top software solution companies across these models before signing. Quote variance of 3 to 5 times for similar work is normal.

A Vendor Worth Considering

If you’re scoping a custom build, Brain Station 23 is one firm that fits this space well. UK presence with offshore engineering depth, operating since 2006, with experience across fintech, retail, manufacturing, and ERP territory. The hybrid onshore-offshore model lines up with where most UK custom build budgets actually land.

The Decision Framework. Cleaner Version

If I had to compress the whole thing into a single decision rule, it’d be this.

Ask yourself two questions. First, is this problem your competitive edge, or operational plumbing? If it’s plumbing, buy. Second, will the SaaS cost over five years exceed the cost of building, owning, and maintaining custom? If yes, build is worth properly considering.

Almost everything else is noise. Most buy-vs-build debates get stuck arguing about features and integrations when the real decision is just those two questions.

A Practical Tip Before You Commit

Before you write a six-figure cheque to anyone, run a paid pilot. Spend £15,000 to £25,000 with one or two shortlisted partners on a small slice of real work. Get them to build something. You’ll learn more about how they operate in 4 weeks than you would in 4 months of sales calls.

Vendors who refuse paid pilots are showing you something. Pay attention to that.

Final Thought

The 2026 build vs buy decision is no longer “buy SaaS unless you have to build.” It’s “buy commodity software, build where you have an edge.” That distinction sounds small but it changes a lot.

For UK businesses sitting between £100,000 and £500,000 of potential build budget, the smart move is shortlisting carefully, running a paid pilot, and being honest about which parts of your business are actually your moat versus which parts are just running costs.

Build the moat. Buy the rest. Most wrong calls come from getting that distinction backwards.