Cost-benefit analysis (CBA) is a method for deciding whether a project is worth doing by putting all of its costs and all of its benefits in the same units, usually dollars over the same time period, and comparing them. On a plant floor, it is the tool that turns "we should fix that" into a one-page case a manager can approve: what it costs, what it returns, how fast, and what happens if we do nothing. Most kaizen events, equipment purchases, safety projects, and Six Sigma charters are justified with one.
When to use it
CBA is for decisions with a real cost and a measurable return, where someone has to choose between doing the project, doing a different project, or doing nothing:
- Buying equipment: a new label printer, a forklift, a replacement press.
- Funding an improvement project: a kaizen event, a 5S rollout, a TPM pilot.
- Safety investments: floor marking, guarding, lighting, mirrors at blind corners.
- Process changes with training and downtime costs attached.
- Choosing between two ways of solving the same problem.
It is not needed for small, reversible changes. Moving a tool rack during a kaizen event does not need a spreadsheet; the decision costs more than the change. CBA earns its time when the project is expensive, hard to undo, or competing with other projects for the same budget.
The six steps
1. Define the project and the alternative
State exactly what will be done and what it is being compared to. The comparison is usually "do nothing," and the costs of doing nothing (the problem continuing) are often the largest number in the analysis. A CBA that forgets to cost the status quo will understate every improvement project.
2. List every cost
- Direct costs. Equipment, materials, software, contractor fees.
- Labor. Hours from the people doing the project, at loaded rates, including training time and the kaizen team's week off their regular jobs.
- Downtime. Production lost during installation, changeover, or the event itself, valued at contribution margin per hour.
- Ongoing costs. Maintenance contracts, consumables, license renewals, the audit time a 5S program needs to sustain.
- Opportunity cost. What the same money and people would have returned on the next-best project. This is what makes CBA a comparison rather than a justification.
3. List every benefit
- Cost reduction. Less scrap, rework, overtime, expediting, inventory, or maintenance. For quality projects, the price of nonconformance framework is the checklist.
- Capacity. More output from the same equipment and people, valued at the margin on the additional units if they can be sold, or at the cost of the capacity you would otherwise have to buy.
- Avoided costs. Injuries, fines, recalls, equipment failures that the project makes less likely. These are probability times consequence, and they are legitimate benefits when the probabilities are honest.
- Time. Hours freed up for operators, maintenance, or supervisors, valued at loaded rate only if the time is actually redeployed.
- Intangibles. Morale, customer confidence, audit readiness. Listed, described, and scored, but kept out of the dollar total (see below).
4. Put numbers on everything over the same period
Pick a time horizon that matches the life of the investment: one to three years for a process change, the equipment's useful life for a machine. Spread costs and benefits across the years they occur. Where a number is uncertain, use a range or a conservative estimate and say so; a CBA with an obviously optimistic benefit line gets rejected, and should.
5. Calculate the comparison
Four measures cover most decisions. Use the one your organization expects, and show the inputs so the reader can check it.
| Measure | Formula | Answers |
|---|---|---|
| Net benefit | Total benefits − Total costs | How much better off are we in dollars? |
| Benefit-cost ratio (BCR) | Total benefits ÷ Total costs | How many dollars back per dollar spent? Above 1.0 is positive. |
| Payback period | Initial cost ÷ Annual net benefit | How long until it has paid for itself? The measure plant managers ask for first. |
| Net present value (NPV) | Sum of each year's (benefit − cost) ÷ (1 + r)year | What is it worth today, given that future dollars are worth less? Use your company's discount rate (r); 8 to 12% is typical. Required for multi-year capital projects. |
6. Test the assumptions
Change the two or three inputs you are least sure of and see whether the answer flips. If the project only works when scrap drops 40% and not when it drops 25%, that is the finding, and the decision should be made knowing it. A sensitivity check takes ten minutes and prevents most bad approvals.
A worked example
Illustrative numbers. A warehouse has had three forklift near-misses at a blind corner in two months, the problem traced in root cause analysis to an unmarked overflow staging area blocking sight lines. The proposed project: define and mark an overflow area, mark a keep-clear zone and stop line at the corner, separate the pedestrian path with tape and signs, and add a convex mirror. The alternative is a safety talk, which costs almost nothing and has already been tried twice.
| Costs (year 1) | |
|---|---|
| Floor tape, corner markers, floor signs, mirror | $1,800 |
| Labor: 2 people × 8 hours at $45 loaded | $720 |
| Downtime: aisle closed 4 hours during a slow shift | $400 |
| Training: 30-minute toolbox talk, 40 people at $45 | $900 |
| Ongoing: tape touch-up and audit time, per year | $600 |
| Total year 1 | $4,420 |
| Benefits (per year) | |
|---|---|
| Avoided forklift-pedestrian injury: 10% annual probability × $60,000 average direct and indirect cost | $6,000 |
| Avoided racking and product damage: 3 strikes per year × $500 | $1,500 |
| Receiving time saved by a defined overflow location: 15 min per day × 250 days × $45 | $2,800 |
| Total per year | $10,300 |
Results. Net benefit in year 1: $10,300 − $4,420 = $5,880. Benefit-cost ratio: 2.3. Payback: $3,820 initial cost ÷ $9,700 annual net benefit (benefits minus the $600 ongoing cost) ≈ 5 months. Over three years at a 10% discount rate, NPV is roughly $20,000. Sensitivity: if the injury probability is 3% instead of 10%, annual benefits fall to $6,100, and payback stretches to about 8 months; the project is still clearly positive. Intangibles, not in the totals: OSHA inspection readiness, operator confidence at the corner, and the fact that the near-miss log now has a documented response.
That last paragraph is the whole deliverable. A manager can read it in a minute, see the assumptions, and approve or push back on a specific number.
Handling intangibles
The standing criticism of CBA is that it ignores what cannot be priced. The practical answer is to keep intangibles out of the dollar total and in a separate, visible list, each one described in a sentence and scored (high, medium, low) for how much it matters. Two projects with similar NPV are then decided by the intangibles, which is where they belong. What not to do is invent a dollar value for morale to make the ratio look better; the first reader who asks where the number came from will discount the whole analysis.
Common mistakes
- Not costing the status quo. "Do nothing" has a cost. For a chronic problem it is usually the biggest line.
- Counting time saved that is not redeployed. Fifteen minutes a day across forty people is a real number only if those hours go to something else.
- Double counting. Reduced scrap and increased capacity from the same process change are often the same benefit measured twice.
- Optimistic single-point estimates. Use ranges or conservative values and show them.
- Ignoring ongoing costs. A 5S program needs audit time forever. A machine needs maintenance. Year 2 is not free.
- Stopping after approval. Measure the actual result at 90 days against the CBA. That is how the next estimate gets better and how the kaizen team keeps its credibility.
Frequently asked questions
What is a good benefit-cost ratio?
Anything above 1.0 means benefits exceed costs. Most organizations set a threshold higher than that, often 1.5 to 2.0, to allow for optimistic estimates and to rank projects against each other.
What is the difference between CBA and ROI?
Return on investment is net benefit divided by cost, expressed as a percentage; a BCR of 2.0 is an ROI of 100%. They are the same comparison in different units. CBA is the broader process of identifying and valuing the inputs; ROI is one of the outputs.
Do I need NPV for a small project?
No. For anything that pays back within a year, payback period and net benefit are enough. NPV matters when costs and benefits are spread over several years, because a dollar in year 4 is worth less than one today.
How do I value an avoided injury?
Multiply the probability of the injury in a year by its expected cost, including direct costs (medical, compensation) and indirect costs (lost time, replacement labor, investigation, equipment damage, insurance). Use your own incident history or your insurer's figures for the cost, and state the probability you assumed so it can be challenged.
What if the benefits are mostly intangible?
Then the decision is not a CBA decision. Present the costs honestly, describe the intangibles clearly, and let the decision-maker decide whether they are worth the price. Dressing an intangible up as a dollar figure does not make the case stronger.
Related tools
CBA is one of the planning tools used before an improvement project starts, alongside the project timeline, resource plan, and stakeholder map, and it is the defense against scope creep once the project is running: every proposed addition gets its own cost and benefit line. For the cost side of quality problems, see Zero Defects and the price of nonconformance; for equipment projects, OEE is where the benefit numbers come from. Harvard Business School Online's overview of cost-benefit analysis covers the general business case method.