Cost Benefit Analysis Template: From Spreadsheet to Sign-Off

Updated May 2026. Rewritten as the cost benefit analysis template we actually run with clients before they sign off an AI or automation build, including the line items most templates forget.

A cost benefit analysis template is meant to do one thing: get a decision approved or killed faster than a meeting could. Most of the templates floating around online produce neither outcome. They produce a spreadsheet with twelve rows of made-up numbers, a positive NPV calculation, and a tab nobody opens after week two.

This guide walks through the cost benefit analysis template we run with mid-market clients before they commit to an AI or automation build. We are a Brisbane automation consultancy, and we have seen this template kill bad ideas in a week and get good ones funded in two. It is structured for finance and IT to both sign off the same document.

It pairs with our reads on calculating ROI on automation, our business process analysis template, and our automation roadmap guide.

What Belongs in a Cost Benefit Analysis Template

Eight sections, in this order. Anything else is decoration.

1. Scope and decision. One sentence on what the decision is and who owns it.
2. Baseline. The current cost or revenue position before any change.
3. Build cost. One-off cost to deliver the change.
4. Run cost. Ongoing monthly or annual cost after delivery.
5. Benefits. Cash, cash-equivalent, and strategic. Kept separate.
6. Risks and assumptions. What has to be true for the benefits to materialise.
7. Payback period and NPV. The two numbers leadership cares about.
8. Decision. Yes, no, or revise. With a date.

Notice what is missing. There is no introduction, no executive summary, no “next steps”. Those belong in a separate memo. The cost benefit analysis template is a working document for finance, not a sales pitch.

Cost Benefit Analysis Template: Section by Section

Scope and decision

One sentence on what is being decided. Bad: “Evaluating AI for accounts payable”. Good: “Whether to build a Claude-based invoice extraction pipeline replacing the manual data-entry team across the AU finance shared service, decision by 30 June, sponsor: CFO.”

If the scope sentence is more than 30 words, the scope is not clear enough. Tighten it before you fill in anything else.

Baseline

What the world looks like if you do nothing. The numbers, not the narrative.

• Annual cost of the current state (salaries, vendor fees, error remediation, opportunity cost).
• Volume metrics (invoices per month, tickets per week, errors per quarter).
• Quality metrics (current error rate, current cycle time, current customer NPS).

If you cannot fill in the baseline, the whole template is built on sand. Spend a week on the baseline before any benefit projection. Most cost benefit analyses we are asked to review fail here.

Build cost

One-off costs to deliver the change. Three buckets, each itemised:

• External. Consultancy fees, software licences, hardware, contractor time. Direct invoices.
• Internal. Hours from your own team. Multiply by fully-loaded cost ($100 to $250 AUD per hour for most knowledge-work salaries).
• Change management. Training, documentation, comms. Usually 12 to 18 percent of the build budget. Most templates forget this entirely.

Run cost

Ongoing cost after launch, monthly. Easy to underestimate.

• Software subscriptions and API costs (Anthropic, OpenAI, vector store, monitoring).
• Infrastructure (hosting, storage, backup).
• Support retainer or internal time to keep it running.
• An annual 15 percent sinking fund against the build cost, to cover the inevitable rework when something breaks or a vendor changes.

The sinking fund is the single most missed line item we see. Software does not stay still for five years. Plan for it.

Benefits

Three categories, kept strictly separate.

Cash benefits. Money that actually leaves or stops leaving the bank account. Vendor cancellations, licence consolidations, contractor hours not bought, revenue uplift backed by a signed pricing change. CFOs respect these.

Cash-equivalent benefits. Hours saved that get redeployed. Real, but only count if there is a plan for the redeployed time. A vague “team focuses on higher-value work” is not a plan. “Three FTEs reallocated to a specific named project” is.

Strategic benefits. Brand, optionality, learning, regulatory positioning. List them, but do not put a number on them and add them to the total. Mixing strategic and cash benefits is how you get a positive NPV that nobody trusts.

Risks and assumptions

Three to seven items. Each one is something that has to be true for the benefits to land.

• Volumes remain at or above current baseline (drop below = ROI shrinks).
• Key sponsor remains in post for 12 months (sponsor turnover kills programs).
• Vendor pricing stable (or capped contractually).
• No regulatory change blocks the implementation.

Each assumption gets a sensitivity number. If volumes drop 30 percent, what does payback look like? If the build runs 25 percent over budget, what does NPV look like? A two-by-two sensitivity grid is enough for most cases.

Payback period and NPV

Two numbers. Both calculated the same way for every project so they compare cleanly.

Payback period. Number of months for cumulative cash benefit to exceed the build cost. We target 8 to 14 months for most automation work. Anything beyond 24 months is treated as strategic, not operational.

NPV over three years. Net present value of cash and cash-equivalent benefits minus build and run costs, discounted at the corporate hurdle rate (usually 8 to 12 percent for mid-market Australian businesses).

A Worked Cost Benefit Analysis Template Example

Real numbers from a finance-team AP automation we shipped last year, anonymised.

Scope. Replace manual data entry on 12,000 invoices per month with a Claude Sonnet 4.5 extraction pipeline plus a validation step, decision by end Q2.

Baseline. 4.2 FTEs on invoice entry at $85,000 AUD fully loaded each, totalling $357,000 per year. Average error rate 1.8 percent, with $48,000 per year in supplier credit memo rework. Cycle time 36 hours from invoice receipt to ERP entry.

Build cost. $48,000 AUD external (consultancy) plus $22,000 internal time plus $11,000 change management. Total: $81,000 AUD.

Run cost. $1,400 AUD per month in Anthropic API, $200 AUD per month hosting, $400 AUD per month observability, plus a $12,000 annual sinking fund. Total: $36,000 AUD per year.

Benefits.

• Cash: 2.5 FTE roles not refilled after attrition, saving $212,500 AUD per year. Credit-memo rework drops to $9,000, saving $39,000.
• Cash-equivalent: remaining 1.7 FTEs reallocated to vendor master maintenance and exception handling, named project, $144,500 worth of hours.
• Strategic: faster month-end close (cycle time drops from 36 hours to 4 hours), improved supplier relationships, eval-pipeline-ready for future AP-adjacent work. Listed, not added.

Risks. Invoice volume stable at 12K per month (sensitivity: at 8K, payback extends from 4 to 7 months). Anthropic pricing stable (contracted ceiling). Sponsor remains in post (CFO confirmed 12-month commitment). Tax authority does not change e-invoicing rules (low risk, monitored).

Payback. $81,000 build divided by ($251,500 annual cash benefit minus $36,000 run cost) = 4.5 months.

NPV over 3 years. ~$540,000 AUD at a 10 percent discount rate.

Decision. Approved, 18 April. CFO sponsor, IT director co-sponsor, AP manager operational owner.

Common Mistakes with Cost Benefit Analysis Templates

• Counting time-savings as cash without a plan to recapture the time. The single most common mistake. If the saved hours go back to the same team doing the same work, the savings are zero.
• Excluding change management cost. Adds 12 to 18 percent to build cost. Programs that skip this on the spreadsheet skip it in execution and underdeliver.
• Forgetting the sinking fund. Year-three rework is not optional. Budget for it.
• One number, no sensitivity. A single-point estimate hides risk. The two-by-two sensitivity grid takes 20 minutes and saves 20 hours of meeting time later.
• Mixing strategic and cash benefits. Inflates the NPV and undermines credibility. Keep them separate, present them separately.

When a Cost Benefit Analysis Template Is Not the Right Tool

Three cases where a cost benefit analysis is the wrong document.

• Compliance work. If you have to do it to meet APP, APRA CPS 234, or a contractual obligation, the question is “do we comply” not “what is the NPV”. Risk register, not CBA.
• Customer-experience work where the metric is qualitative. A redesign of an onboarding flow does not have a clean cash number. Use a target metric (NPS, completion rate, time-to-first-value) and a budget envelope instead.
• Strategic options. “Should we enter the US market” is not a CBA question on a one-page template. It is a multi-year scenario model.

Where the cost benefit analysis template earns its place: process automation, software replacement, AI builds, system integrations, vendor consolidations. Anything with a measurable baseline, a finite build budget, and a 3-to-5-year horizon.

If you would like us to run our cost benefit analysis template on a specific automation or AI build, send us a note via the contact page or book a call. We do this for free on shortlisted projects because the conversation is shorter when the numbers are on the table.

Frequently Asked Questions

What is a cost benefit analysis template?

A cost benefit analysis template is a structured spreadsheet (or one-page memo) that compares the build and run costs of a project against the benefits, ending in a payback period and NPV. Used to decide whether to fund or kill a proposed initiative.

What is the difference between a CBA template and an ROI calculation?

ROI is a single number (return as a percentage of investment). A cost benefit analysis template is the full working that produces that number, plus payback period, NPV, sensitivities, and risk assumptions. ROI is the output. CBA is the document.

What makes a good benefit-cost ratio?

Any benefit-cost ratio above 1.0 means benefits exceed costs. In practice we look for 2.0 or higher on automation work, which translates to a payback period under 14 months. Anything below 1.5 is treated as marginal and requires stronger strategic justification.

How do you value intangible benefits in a cost benefit analysis template?

List them, do not number them. Strategic, brand, optionality, and morale benefits matter but should not be added to the cash total. Reporting them separately preserves the credibility of the cash NPV. CFOs see straight through projects where intangibles inflate the headline number.

What cost categories do CBA templates usually miss?

Change management (training, documentation, comms), the sinking fund for year-two and year-three rework, internal time at fully-loaded cost, and vendor price increases over the 3-to-5-year horizon. These four together routinely add 25 to 40 percent to the all-in cost.

How far should a cost benefit analysis template look ahead?

Three years for most automation and AI work. Long enough to see benefits land. Short enough to keep forecasts honest. Five years for capital-heavy work like ERP replacements. Beyond five years the forecast is fiction.

What format should the cost benefit analysis template be in?

One page. Excel or Google Sheets with a single readable summary tab and detailed working tabs underneath. PowerPoint versions tend to hide assumptions in the speaker notes and get used to sell rather than to decide. The summary tab is the document; everything else is showing the working.

Who should own the cost benefit analysis template?

The business sponsor, not the consultancy proposing the work. If the consultancy writes the CBA, the numbers will be optimistic. The business sponsor writes the CBA with input from finance, IT, and the consultancy. That is the only structure that produces decisions you can later defend.