In plain English: Buy for standard processes, bend the tools you have when they don’t connect, build when the process is where you win. Count the hours people spend bridging gaps.
Picture a perfectly normal company. It pays for five SaaS tools, two integration platforms and one very patient employee whose real job — never written down anywhere — is to copy data from one tool into another.
Management calls this "the cheap option".
It isn't. It's just the option whose costs are hidden in payroll, where nobody looks.
"Don't build what you can buy" is good advice. It's also incomplete. There aren't two options, there are three — and picking the wrong one quietly costs companies tens of thousands of euros a year.
The three options
Buy — choose a standard product and adapt your process to it. You get a mature product, updates, support and a predictable monthly bill. You give up control over how the process works.
Bend — keep the tools you have and make them fit better: configuration, automation rules, integrations between systems, a few small scripts. Often the best value of all — and the option people forget exists.
Build — develop software for your specific process: a client portal, an internal cockpit, a pricing engine, the missing link between two systems. You get exactly what you need and you own it. You also own the maintenance.
When to buy
Buy when the process is standard and not what makes you different. Accounting, payroll, e-mail, file storage, basic CRM — thousands of companies do these the same way, and the products reflect decades of accumulated best practice. Adapting to them is usually an upgrade, not a compromise.
Warning signs that buying is the wrong call:
- Every demo ends with "it does 80 % of what we need".
- The missing 20 % is exactly the part your customers notice.
- You're planning workarounds before you've even signed the contract.
// quick one
Want us to do this for you? We only take a few new clients at a time.
See if you’re a fitWhen to bend
Bend when your tools are basically right but refuse to talk to each other, or when a handful of steps are still manual. Classic examples:
- Syncing won deals from the CRM into the invoicing tool
- Creating a project automatically the moment a contract is signed
- A nightly job that pulls numbers from three systems into one report
Bending is cheap and fast — often days, not weeks. The danger is accumulation: ten automations, built by three different people over four years, undocumented, understood by nobody. If you bend, document every integration and give it an owner.
Bending stops being cheap when:
- More than a couple of people spend real time babysitting workarounds.
- The integrations break every time a vendor ships an update.
- You're paying for several tools just to use one feature of each.
When to build
Build when the process is core to how you win, or when you're paying people to bridge gaps every single day. Strong candidates:
- A client portal that makes you noticeably easier to work with than your competitors
- Internal tools for a workflow no standard product models well — specialised quoting, scheduling, compliance
- A layer that ties several systems together and ends manual reconciliation for good
And no, custom software doesn't have to mean a two-year saga. A focused tool that does one job well can often be in production in six to ten weeks, then grow from there.
The maths nobody does
Compare the options over three years, not one. For each, add up:
| Cost | Buy | Bend | Build |
|---|---|---|---|
| Licences / subscriptions | High, per user, rising | Existing tools + integration platform | Low (hosting) |
| Implementation | Low–medium | Low | Medium–high |
| Ongoing maintenance | Included | Grows with every integration | Planned — a common rule of thumb is 15–20 % of the build cost per year |
| People bridging the gaps | Depends on fit | Usually drops | Usually drops most |
| Fit with your process | You adapt | Mostly fits | Exact |
The row almost every comparison forgets is people bridging the gaps. Say two employees each spend five hours a week moving data between tools. That's over 500 hours a year. At a loaded cost of €50 an hour, that's more than €26,000 annually — often more than the software bill itself. (Illustrative numbers — run your own with the calculator on our homepage.)
The decision rule (print it, stick it on the wall)
- A standard product fits with only minor compromises → buy.
- Your tools fit but don't connect, or a few steps are manual → bend — and document it.
- The process is where you're different, or the workarounds cost more than a focused tool → build something small, then grow it.
Then revisit the decision every couple of years. Yesterday's sensible "bend" can be today's expensive tangle — and the tool you built may now exist off the shelf.
Not sure which bucket you're in? The free Digital Readiness Scorecard sizes up your software landscape and four other areas in three minutes. If a project makes sense, you can book a strategy call straight from your result.