Build vs buy: sales systems for service firms

The short answer: buy off-the-shelf software when your problem is generic, build a system when the problem is your specific pipeline, and hire people only when judgement is the actual bottleneck. Most 5–50 person service firms get this backwards — they hire a £35k+/year BDR to do work a system handles better, while buying generic tools that solve problems they never had.

I build these systems for a living, so discount my bias accordingly. But the framework below is the one I use on my own firm, and it holds up whether or not you ever buy from me. The wider context — what automation genuinely delivers versus what the vendors claim — is in AI Automation for B2B: what actually works.

What do "build" and "buy" actually mean here?

"Buy" means subscribing to software that does the job as-is: a CRM, an email verification service, a scheduling tool. "Build" means assembling components into a system shaped around your process — your ICP, your sequences, your handover points — whether you assemble it yourself or pay a specialist to. "Hire" is the third option nobody frames as a build-vs-buy decision, though it is the most expensive one on the menu.

The distinction that matters is not technical. It is this: bought software makes you adapt to it; a built system adapts to you. Both are fine. Problems start when you buy something generic and then spend a year forcing it to fit, or build something custom for a problem a £30/month subscription already solves.

When is buying off-the-shelf the right call?

Buy when the problem is commodity. Calendar booking, e-signatures, invoice chasing, transcription — these are solved problems, and the vendors have refined them across thousands of customers. Building your own version is vanity engineering.

The test I use is a simple mechanism. When a process is identical across most firms in your category, then buy the tool and adopt its defaults. When a process is specific to how you win work — who you target, what you say, how leads are routed and followed up — then a generic tool will only ever be a container, and the system around it has to be built. When the process needs discretion on every single case — pricing a complex engagement, handling a delicate renewal — then neither software nor system replaces a person, and you hire.

Run each candidate process through those three gates in order and the build-vs-buy question mostly answers itself.

What does building actually cost?

Concrete numbers, since this is where hype usually replaces arithmetic. Our Outbound Engine — list building, enrichment, verification, warmed sending infrastructure, sequences, reply handling — runs £4,000–£6,500 as a one-off build, or £1,500–£3,000/month fully managed, live in 30 days. A standalone verified prospect database is £950.

Compare the default alternative: a BDR at £35k+/year — £2,900+ a month before management time, holidays, ramp-up, and the risk they leave with the process in their head. The system costs less, does not resign, and its output is owned by the firm rather than the employee. That ownership point matters more than the price gap: when growth depends on individuals rather than assets, you have a fragility problem, and these five signs your growth depends entirely on you show how it compounds.

None of this makes building automatically right. A built system needs maintenance, and someone accountable for it. Workflows rot like everything else.

Should you build it yourself or pay someone?

If you have a technically curious operator with genuine slack in their week, building in-house on n8n or Make is viable and instructive. The tools are accessible; the discipline is the hard part. What typically goes wrong is not the first build but the eighteenth month — undocumented workflows, one person who understands them, and a business that cannot touch its own plumbing when that person is on holiday.

If you pay a specialist, buy a fixed scope with defined deliverables and keep ownership of every account and credential. The warning signs of a bad vendor are consistent enough that I wrote them up separately in red flags when hiring an AI automation agency — tools before process, vague scope, and ownership traps head the list.

Where should a service firm start?

Not with the most impressive automation — with the most expensive gap. For most firms of this size that is either the empty top of the funnel (no consistent outbound, pipeline hostage to referrals) or the leaky middle (enquiries answered slowly, follow-up stopping after two touches when deals typically need five or more). Both are system problems, not effort problems.

Before building or buying anything, be clear about what AI can actually automate in a service firm and what still needs a human. Then price all three options — buy, build, hire — against the same gap, in pounds per year. The spreadsheet is rarely kind to the default choice.

---

Next step: the Growth System Audit — £450, seven days, credited against any build — maps where your growth system leaks and what to build first.