Strategy & Leadership

Lesson 5 · Finance

Making the business case

Translate a technical proposal into the language of the people who fund it.

A business case translates a technical proposal into the language of the people who fund it: what the problem is, what solving it is worth, what it costs, what could go wrong, and what you are asking for. One page beats twenty slides.

Quantify benefits in one of three currencies: revenue gained, cost saved or risk reduced. Engineering time saved only counts if it is redeployed to something valuable, so say what. For risk, use expected value: probability × impact. A 10% annual chance of a $2M outage is a $200,000-a-year expected cost.

Then stress-test it. Sensitivity analysis asks which assumption, if wrong, kills the case. Present a base, best and worst case, and name the leading indicator that will tell you early if it is off track. Finally, compare against the alternatives, including doing nothing — which always has a cost of its own, and stating that cost is often the strongest argument.

Picture it

“We need to pay down tech debt” gets no budget. The rewrite: “Our deploy pipeline fails one release in five. With 250 releases a year, that is 50 failures at about 20 engineer-hours each — roughly $100,000 a year. The fix is two engineers for six weeks, about $48,000. Payback is under six months. The main risk is the migration overrunning; we will know by week three.” Same project. Funded in one meeting.

Key formulas & frameworks

Business case
Problem → Options → Costs → Benefits → Risks → Ask
Benefit currencies
Revenue gained · Cost saved · Risk reduced
Expected value
Probability × Impact
Sensitivity test
Which assumption, if wrong, flips the decision?

Try it — NPV & payback

Net present value

$23,291

Undiscounted ROI

30%

Payback

2.3 yr

PV of benefits

$323,291

Now
−$300,000
Yr 1
$118,182
Yr 2
$107,438
Yr 3
$97,671

Each bar is a cash flow in today's dollars. Push the start date out a year or raise the discount rate and watch a project that looked comfortable turn negative.

Practice

Score: 0 / 5

Question 1

Q1.There is a 5% annual chance of a breach that would cost $4M. What is the expected annual cost, in dollars?

Question 2

Q2.How should the “do nothing” option be treated in a business case?

Question 3

Q3.Flaky tests waste six hours per engineer per month. With 25 engineers at $100 an hour, what is the annual cost, in dollars?

Question 4

Q4.Engineering hours saved always translate directly into money saved for the company.

Question 5

Q5.What is sensitivity analysis for?

Up nextDecision-making

Next lesson