Every growing organization hits this fork eventually. A process that used to run on a spreadsheet and a shared drive starts breaking under its own weight, and someone asks the obvious question: do we buy something built for a thousand other companies, or do we build something built for us? Both answers are defensible. Both are also wrong for most of the organizations we've watched make this call badly.
Here's a framework that skips the consultant hedge of "it depends" and gets to an actual answer.
Start from the default: buy
BDC's guidance to small business owners on this is blunt: out-of-the-box solutions are almost always cheaper in the long run, more scalable, and better supported than most custom-built alternatives (BDC). The hard part isn't accepting that advice. It's finding the right platform among the dozen that claim to fit.
There's a second reason buy is the right default for most Atlantic Canadian SMBs specifically: most haven't finished digesting the tools they already own. CFIB found that only about a tenth of Canadian small businesses have fully integrated digital tools across their operations, even though the ones who do see real returns, roughly $1.60 back for every dollar spent and a 29 percent average productivity gain (CFIB, 2025). If your organization is still at that stage (tools half-adopted, workflows half-migrated) a custom build is almost never the right next move. You'd be pouring a foundation under a house you haven't finished furnishing.
When custom actually wins
There are real cases, and they cluster around three patterns.
- The process is the product. If how you do a thing is the actual reason customers choose you over a competitor, forcing that process into someone else's generic workflow flattens the advantage. A fabricator with a scheduling logic tuned to how their specific shop floor actually moves shouldn't bend that logic to fit a $40-a-seat SaaS tool built for a generic trades business.
- No off-the-shelf option covers the actual workflow, and you've genuinely checked. Not "we looked at two tools and neither had the exact button we wanted." Checked, as in six vendor demos, a written list of dealbreakers, and an honest answer about whether a 20 percent workflow change on your end would make the ninth option fit.
- Integration costs have quietly become the real price tag. This is the one people miss. A platform that looks cheap at $200 a month can cost five figures a year once you count the middleware, the API calls to your existing systems, and the staff hours spent working around what it can't do.
If none of those three genuinely apply, buy. Most of the time, none of those three genuinely apply.
What getting it wrong actually costs
McKinsey and the University of Oxford studied more than 5,400 large IT programs (initial cost estimates above $15 million, so a different scale than most Atlantic Canadian SMB projects) and found that on average they ran 45 percent over budget while delivering 56 percent less value than planned (McKinsey and University of Oxford, 2012). The dollar figures in that study don't map onto a 30-person Moncton distributor deciding between a $15,000 inventory build and a $6,000-a-year off-the-shelf plan. The pattern underneath it does. Scope creeps once real users touch the thing. Integration takes longer than the estimate. The team that was supposed to maintain it moves on. Those forces don't care whether the project is worth $15 million or $15,000. They just show up smaller.
The stakes are real either way. BDC's research puts the average Canadian SMB's annual digital technology investment at roughly $118,000 (BDC, 2022), and the same study names the barriers that trip organizations up regardless of which path they pick: cost (42 percent), cybersecurity (32 percent), uncertain benefits (27 percent), and integration with existing systems (27 percent). Buy doesn't dodge those risks. It just spreads them across a vendor's existing customer base instead of concentrating them on your own build team.
Five questions before you decide
Answer these honestly, in writing, before a vendor demo or a developer quote enters the room.
- Does this process make us money differently than our competitors, or does it just run the business? Differently: lean build. Just runs the business: lean buy.
- What's the real three-to-five-year cost, not the year-one quote? Licensing compounds. So does maintenance.
- Who maintains this in year two if the person who built or configured it leaves? A vendor's support line has an answer. A single contractor who built your custom tool and then took a job in Toronto does not.
- What happens if the vendor gets acquired, changes pricing, or shuts down? Real question for a buy decision, rarely asked.
- Do we actually have six months for a build, or do we need something running in three weeks? Timeline pressure alone answers this one most of the time.
The honest recommendation
For an Atlantic Canadian SMB under a few hundred people, buy is the right call roughly four times out of five. That's not a hedge, it's an opinion, and it's the opposite of what a lot of build-focused shops will tell you, because a lot of build-focused shops make their money on the fifth case. We don't take vendor commissions either way, which is the only reason we can say this without an angle: the honest starting posture is skepticism toward custom work, not enthusiasm for it.
The 20 percent where build wins is usually obvious once you stop rationalizing. It's the workflow that actually differentiates you, not the plumbing around it. If you're not sure which one you're looking at, that uncertainty is itself useful information. It usually means you're looking at plumbing.
This is the kind of call we help structure under Architecture & Vendor Selection, Stage 6 of our methodology: a scored build-vs-buy analysis, a real vendor short-list, and a reference architecture that an engineer can actually build from. We've also written about what this kind of consulting actually costs, and about the grant programs that can fund either path, which is worth knowing before you assume build is unaffordable or buy is free.
If you're stuck on this decision right now, a free 30-minute call is enough time to walk through your specific situation and get a straight answer, not a sales pitch for whichever path we happen to sell.