Step 22 of 23
where money goes, hidden costs, prioritization framework, common mistakes
เงินไปไหนบ้าง — คน 50-70%, Infra 15-25%, License 10-20%, ซ่อมบำรุง 10-20%
Technology spending falls into four categories. Most organizations underestimate at least two of them:
| Category | What It Covers | Typical Split |
|---|---|---|
| People | Salaries, benefits, training, recruiting | 50-70% of total tech budget |
| Infrastructure | Cloud services, hosting, tools, hardware | 15-25% |
| Licenses and SaaS | Software subscriptions, third-party services | 10-20% |
| Maintenance and Operations | Bug fixes, updates, security patches, monitoring | 10-20% |
The people line is almost always the largest. This is why "we can just hire more engineers" is an expensive solution -- each hire costs their salary plus benefits, plus tools, plus management overhead, plus onboarding time (3-6 months before full productivity).
Budgets often miss costs that are real but not obvious:
Training and onboarding. New tools require training. New hires require onboarding. Estimate 2-3 months of reduced productivity for every new team member or new system.
Migration. Moving from one system to another is expensive: data migration, re-training, parallel running, and the inevitable disruption. Budget 30-50% of the original implementation cost for migration.
Opportunity cost. Every hour your team spends maintaining an old system is an hour not spent building something new. This is the hardest cost to quantify and the most important.
Technical debt interest. Shortcuts taken earlier consume ongoing engineering time. If your team spends 30% of their time working around old shortcuts, that is 30% of your people budget spent on debt service.
Compliance and security. Audits, certifications, penetration testing, and compliance tooling are not optional in regulated industries. They are recurring costs.
When budget is constrained (it always is), use this priority order:
| Priority | Category | Rationale |
|---|---|---|
| 1 | Security and compliance | Non-negotiable. Breach costs exceed prevention costs. |
| 2 | Maintenance and reliability | Keeping existing systems running. Cannot build new things if current things are broken. |
| 3 | Revenue-generating features | Features that directly drive sales, retention, or efficiency. |
| 4 | Operational efficiency | Tools and automation that reduce manual work and save time. |
| 5 | Strategic investments | Long-term bets (new platforms, AI experiments, rewrites). High uncertainty, high potential. |
Security and maintenance are never the exciting budget items. They are the ones that prevent disasters. Underfund them and every other investment is at risk.
Budgeting for build, not for run. It costs money to keep software running after it is built. Ongoing costs are often 40-60% of the initial build cost per year.
Annual budgets for monthly decisions. Cloud costs fluctuate with usage. SaaS pricing changes at renewal. Budget flexibility matters.
Ignoring shadow IT. Departments buying their own tools outside of IT's budget. Real spending is often 30-50% higher than the official IT budget because of these untracked purchases.
A well-structured tech budget is transparent about the full cost of technology, honest about the trade-offs, and realistic about hidden expenses.
When reviewing a tech budget, ask: "Does this cover maintenance, training, and contingency? Or only new features?" If the answer is the latter, you are budgeting for a future crisis.
