Buy vs. Build: A Framework for Smarter Decisions
A decision framework for when to buy, when to build, and how to avoid expensive mistakes.
Choosing between off-the-shelf software and a custom build is one of the highest-stakes technology decisions a business makes. Get it right and you move fast, spend efficiently, and stay focused on what makes you competitive. Get it wrong and you spend 3–10x more than expected, delay operations by months, and end up maintaining a system that was never quite right.
This framework cuts through the noise.
First, Define Your Terms
Off-the-shelf software (also called commercial off-the-shelf, or COTS) is pre-built software sold to many customers. Think QuickBooks, Salesforce, SAP, or any SaaS product you pay a subscription for. It is designed to handle common business needs and is ready to deploy in days or weeks.
Custom software is built specifically for your organization, from the ground up, to match your exact processes, data structures, and workflows. It takes months to build, requires ongoing maintenance, and only you own it.
Neither is inherently better. The right answer depends entirely on your context.
The 5-Question Framework
Before writing a single line of code or signing a vendor SLA, answer these five questions honestly.
1. Is this a competitive advantage?
- Yes (Core differentiator): A unique workflow, proprietary process, or capability that competitors cannot easily replicate. Build.
- No (Standard business function): Payroll, basic CRM, email, file storage. Buy. Everyone does this the same way, and that is fine.
2. What is your real timeline?
Need something working within three months? Buy. Custom software realistically takes six to eighteen months from discovery to deployment — and that is for a well-run project with clear requirements. If you have external deadlines, compliance dates, or a market window to hit, off-the-shelf is the only sane choice.
3. Can you afford total ownership?
- Off-the-shelf: Predictable subscription cost. Vendor handles infrastructure, security patches, and feature updates.
- Custom: Build cost, plus roughly 20% of that cost every year in maintenance, bug fixes, documentation, and onboarding new developers who need to understand the codebase.
A system that costs $50,000 to build costs $10,000/year to maintain indefinitely. Factor this in before you start.
4. Does your team have the expertise to own this?
Custom software without a dedicated product manager and at least one senior architect on the client side almost always fails — not because the code is bad, but because requirements drift, priorities shift, and no one is accountable for the product decisions that need to be made every week.
No internal product leadership? Do not build custom. You will regret it.
5. How stable are your requirements?
- Changing every quarter: Buy. Off-the-shelf vendors absorb market changes into their product. You benefit without rebuilding.
- Frozen and well-understood for two or more years: Custom may be worth it. Stable requirements are the single biggest predictor of a successful custom build.
The Golden Rule
Buy commodity. Build differentiators. Integrate everything else.
This is not a compromise. It is a strategy. The best technology organizations in the world buy aggressively for anything that is not their core business, and build only where they can win.
When to Buy
- Standard business functions: accounting, HR, payroll, file storage, basic project management
- You are paying for features you do not use — that is fine, because you are also paying for security, compliance, and a team of engineers maintaining it
- Compliance is the vendor's problem (SOC 2, HIPAA, GDPR) — this alone can justify the subscription cost
- You are early stage and your processes are still evolving
Rule of thumb: if ten other companies in your industry use the same tool, buy it.
When to Build
- The software is your product — you are a SaaS company or technology platform
- Off-the-shelf forces dangerous workarounds: manual re-entry between systems, shadow spreadsheets, staff inventing unofficial processes
- You have dedicated product leadership and a 12+ month budget
- Your workflow is genuinely unique — not "we do things slightly differently," but fundamentally incompatible with any existing tool
The Dangerous Middle (Where Money Dies)
Trap 1: "We are special, so we need to build." Most organizations are not as operationally unique as they think. Eighty percent of companies that build custom software could have bought something, configured it well, and moved faster. Challenge your team: have you actually exhausted the market?
Trap 2: "We bought it and now we are hacking it to death." If your team has built more than 20% custom functionality on top of a commercial product, you are now maintaining a custom system anyway — one with the worst of both worlds. Stop. Evaluate whether you should migrate to a proper custom build or a different off-the-shelf product.
Decision Matrix
| Your Situation | Verdict |
|---|---|
| Startup, pre-product-market fit | Buy everything. Iterate fast. |
| Scale-up with a truly unique core workflow | Build core. Buy the rest. |
| Enterprise with 50+ Salesforce customizations | You built already — just poorly. Evaluate migration. |
| "No tool can do what we need" | 90% chance you have not looked hard enough. |
| Requirements change every quarter | Buy. No debate. |
| Software is your product | Build. That is your job. |
The Bottom Line
When in doubt, buy. Most custom software projects fail not because of bad engineering, but because requirements change faster than delivery. The graveyard of failed software projects is full of systems that were beautifully built for a problem that no longer existed by the time they launched.
Save your build budget for what makes you genuinely different. Buy everything else, and buy it fast.
