Budgets, ROI & Vendors
Managing the money. Learn how to prove the financial value of quality to the business.
1 The Hook
A Test Manager at a large NZ utility was asked to cut their budget by 15%. Instead of just cutting people, they presented a **Cost of Quality (CoQ)** report. They showed that spending $500k on automated regression was saving $2M per year in manual overtime and production hotfixes.
In the C-Suite, quality is a line item. You must learn to speak in dollars.
2 The Rule
Good testing doesn't cost money; it saves it. Proving ROI is the difference between being a "Cost Centre" and a "Strategic Partner."
Senior engineer insight
The moment that changed how I think about QA budgets was watching a CFO reject a request for two additional testers — then approve a $400k emergency vendor contract three months later to clean up a production data corruption incident that those testers would have caught. The lesson: you are not budgeting for testers, you are pricing risk transfer. Once I started framing every budget line as "the premium on an insurance policy against a specific failure mode," budget conversations became dramatically easier.
The most common mistake Test Managers make is defending their budget with effort metrics (hours run, test cases executed) rather than failure cost data — give a CFO a number for what a production outage costs per hour, and your automation investment justifies itself immediately.
From the field
On a large NZ government agency modernisation programme — the kind that runs across two budget years and involves Treasury sign-off mid-delivery — the QA team had their headcount frozen in the second-year budget because "the platform is built, testing is done." The assumption was that testing is a phase, not a function. What the agency discovered when they deferred regression on the tax calculation module was a $1.1M overpayment error that hit 4,200 benefit recipients — and the remediation cost dwarfed the frozen headcount. The team rebuilt their business case using CoQ data from that incident and secured a permanent QA line item in the next annual appropriation cycle. The lesson that generalises: in NZ public sector, budget cycles are fixed and political, so you need failure-cost evidence from inside the organisation — not industry benchmarks — to survive a freeze.
3 Watch Me Do It: The "Root Canal" Bill
Observe how we categorise testing costs to show stakeholders the "Savings" vs. the "Spend." This is how you defend your budget.
| Category | The "Spend" (Appraisal) | The "Savings" (Internal Failure Avoided) |
|---|---|---|
| Automation | $50k (Tool + Setup) | $200k (2,000 manual hours @ $100/hr saved) |
| Shift-Left Reviews | $10k (Staff time) | $80k (Avoiding 4 weeks of rework in dev) |
| Security Scan | $5k (Licence) | $500k (Avoiding a data breach disaster) |
Manager Insight: By presenting the "Savings" column, you change the conversation from "How much do we pay?" to "How much risk are we removing?"
4 Budget Lab: The Cut
In this lab, you must defend your testing function against a 10% budget cut. Use the **Cost of Quality (CoQ)** framework.
Your Task: The 10% Defense
The CFO wants to save $100k. They suggest removing the "Managed Service" (offshore testers) for 3 months.
WORK THROUGH THESE STEPS:
- Impact Mapping: Which projects will stop? What is the daily "Cost of Delay" for those projects?
- Risk Analysis: If regression testing isn't done, what is the "External Failure Cost" (Production bugs)?
- Alternative: Can you save $100k elsewhere (e.g. tool consolidation) without stopping execution?
Draft a 3-bullet point email to the CFO. Focus on "Risk" and "TCO," not "Hard Work." Write it in your notes.
5 Common Mistakes
⚠ Hiring on "Hourly Rate" alone
Why it fails: A $20/hr junior often has a higher **TCO (Total Cost of Ownership)** than an $80/hr senior due to the time you spend managing them and the rework required.
⚠ Budgeting by "Headcount" instead of "Risk"
Why it fails: If a high-risk project gets 1 tester and a low-risk project gets 1 tester, you are wasting money. Allocate dollars where the **Failure Cost** is highest.
Why teams fail here
- Presenting testing costs without the corresponding "failure avoided" column — you look like a cost centre, not a risk function, and the CFO has no reason not to cut you.
- Budgeting by headcount parity ("last year we had six testers, this year we need six testers") instead of by project risk profile — risk-blind allocations waste money on low-risk work and under-resource high-stakes releases.
- Treating vendor TCO as the invoice amount — ignoring knowledge transfer costs, context ramp-up time, management overhead, and the hidden cost of defect leakage from teams unfamiliar with your domain.
- Waiting until the annual planning cycle to make the ROI case — by then decisions are already locked; Test Managers who win budget build the evidence base continuously throughout the year from incident data, escaped defect logs, and automation savings reports.
Key takeaway
A QA budget defended with failure-cost data is nearly impossible to cut — a QA budget defended with effort metrics is the first thing that goes.
6 Self-Check
Q1. What is the difference between Appraisal Costs and Failure Costs?
**Appraisal** is what you spend to find bugs (Testing). **Failure** is what it costs when you don't find them (Rework/Production Outages).
Q2. How do you prove ROI for a testing tool?
Compare the (Cost of Manual Execution over 12 months) vs. the (Cost of Tool + Script Maintenance over 12 months).