Purpose
Turn raw operational inputs — labor cost, process volume, error rates, and automation investment — into a defensible financial business case with ROI, payback period, NPV, IRR, and a sensitivity analysis that shows which assumptions actually move the decision. Solves the problem of automation projects being pitched on gut feel rather than numbers finance will approve.
When to Use
Invoke this skill when the user asks to:
- Justify or build a business case for a process automation investment
- Calculate ROI or payback period for RPA, workflow tools, or process automation
- Compare the financial return of automating vs. keeping a manual process
- Model NPV or IRR for an automation project
- Present automation savings to a CFO, finance team, or leadership
- Stress-test whether an automation project's numbers hold up under conservative assumptions
Inputs Required
Before starting, collect:
- Process description: What the manual process is and the tasks being automated
- Labor cost: Fully-loaded hourly cost of staff doing the work (or annual salary + burden rate)
- Time per transaction: Minutes the manual process takes per unit/transaction
- Process volume: Transactions per month or year
- Error rate & rework cost: Current error percentage and average cost to correct each error (rework time or downstream impact)
- Automation investment: One-time implementation cost (software, integration, setup, training)
- Ongoing costs: Annual license, maintenance, or support fees for the automation
- Automation effectiveness: Expected % of manual work eliminated and residual error rate after automation (default: 90% of work eliminated, error rate reduced by 80% if not provided)
If discount rate is not provided, default to 10% for NPV. If analysis horizon is not specified, default to 3 years. State any defaults used.
Instructions
Step 1: Establish the Manual Process Baseline (Annual Cost)
Calculate what the process costs today, before automation.
- Annual labor cost = (time per transaction in hours) × (annual volume) × (fully-loaded hourly cost)
- If salary was given instead of hourly rate: hourly cost = annual salary × burden multiplier (default 1.3 for benefits/overhead) ÷ 2,080 working hours
- Annual error cost = (annual volume) × (current error rate) × (cost per error correction)
- Total current annual cost = annual labor cost + annual error cost
Show each figure with the formula and the plugged-in numbers so the buyer can audit the math.
Step 2: Model the Automated Process (Post-Automation Cost)
Calculate residual cost once automation is running.
- Residual labor cost = annual labor cost × (1 − % of work eliminated) — accounts for exceptions, oversight, and edge cases still handled manually
- Residual error cost = annual volume × (post-automation error rate) × cost per error
- Ongoing automation cost = annual license/maintenance/support
- Total automated annual cost = residual labor + residual error + ongoing automation cost
Step 3: Calculate Annual Savings and Core Returns
- Annual net savings = total current annual cost − total automated annual cost
- ROI (Year 1) = (annual net savings − automation investment) ÷ automation investment × 100
- Simple ROI (steady state) = annual net savings ÷ (automation investment + first-year ongoing cost) × 100
- Payback period = automation investment ÷ annual net savings (express in months: × 12)
Step 4: Multi-Year NPV and IRR
Model the investment over the analysis horizon (default 3 years).
- Year 0: −(automation investment)
- Years 1–N: +annual net savings (net of ongoing costs, already accounted in Step 2)
- NPV = Σ [cash flow in year t ÷ (1 + discount rate)^t] for t = 0 to N
- IRR: the discount rate at which NPV = 0 — approximate and state it
- Present a year-by-year cash flow table: Year | Cash Flow | Discount Factor | Present Value | Cumulative PV
Step 5: Sensitivity Analysis
Stress-test the three assumptions most likely to be wrong, showing outcome under each scenario.
Vary each of these ±25% (or to a realistic conservative bound) while holding others constant:
- Process volume — what if throughput is lower than projected?
- Automation effectiveness (% of work eliminated) — what if it only eliminates 70% instead of 90%?
- Labor cost — what if fully-loaded cost was overstated?
For each variable, present a small table: Scenario (Low / Base / High) → resulting Annual Savings, Payback (months), 3-Year NPV. Then identify which single assumption the business case is most sensitive to.
Step 6: Verdict and Recommendation
State a clear conclusion:
- Whether the investment clears a reasonable hurdle (positive NPV, payback under 18 months is generally strong for automation)
- The break-even condition — the point at which the project stops being worth it (e.g., "remains positive as long as volume stays above X or effectiveness stays above Y%")
- One risk or caveat the buyer should flag to finance
Output Format
- Structure: Sequential sections following the six steps; financial figures shown with formulas and plugged-in numbers; cash flow and sensitivity presented as tables
- Length: 700–1,000 words plus tables
- Format: Structured financial memo — narrative for interpretation, tables for numbers
- Sections: Manual Baseline → Automated Process Cost → Annual Savings & ROI/Payback → 3-Year NPV & IRR (cash flow table) → Sensitivity Analysis (3 tables) → Verdict & Recommendation
- Tone: Finance-literate and direct — the tone of an operations analyst presenting to a CFO. No hype about the automation itself; let the numbers argue.
Quality Checklist
Before delivering output, verify:
- [ ] Every core figure (labor cost, error cost, savings, ROI, payback, NPV) shows its formula with the actual numbers plugged in
- [ ] Residual labor and residual error costs are subtracted — automation is not modeled as eliminating 100% of cost
- [ ] The NPV calculation includes the Year 0 negative outflow and discounts each future year correctly
- [ ] Payback period is expressed in months
- [ ] Sensitivity analysis covers all three variables with Low/Base/High scenarios in table form
- [ ] The recommendation identifies the single most sensitive assumption and states a break-even condition
- [ ] Every default assumption used (discount rate, burden multiplier, effectiveness, horizon) is explicitly stated
Notes
- The most common overstatement in automation business cases is assuming 100% labor elimination — always model residual manual effort for exceptions and oversight, or the numbers won't survive finance scrutiny.
- Error-cost savings are frequently the hidden driver of ROI; if the buyer omits error/rework cost, prompt for it, since automation often justifies itself on quality gains more than raw labor.
- If payback exceeds the analysis horizon or NPV is negative, say so plainly rather than reframing — a business case that doesn't hold up is more valuable to surface early.
- IRR here is an approximation; note that a spreadsheet or finance tool should confirm it before formal submission.
How to use this skill
This file is a Claude Code skill in Anthropic's standard format — markdown with YAML frontmatter at the top.
Installing in Claude Code:
~/.claude/skills/<folder-name>/SKILL.md— auto-loads in every session anywhere on your machine (recommended for skills you'll use across multiple projects)<your-project>/.claude/skills/<folder-name>/SKILL.md— auto-loads only when working in that project (recommended for project-specific skills)
Restart Claude Code or open a new session to register.
Using in Claude.ai, ChatGPT, Gemini, or other AI tools:
- Drag this file into a new chat, or copy and paste its contents
- Send your task or request as your next message — the AI will apply the skill where it fits
For Claude Code skill details: https://docs.anthropic.com/en/docs/claude-code/skills