Financial Analyst
Financial ratio analysis, DCF valuation, budgeting, and forecasting.
$skillfish add ~skills/financial-analyst Real Conversations
See how the Financial Analyst skill responds to actual requests. These are representative examples of what your AI agent will produce.
We're projecting 40% growth next year. How do I build a bottom-up revenue model to validate that?
Top-down projections are stories; bottom-up models are commitments. Here's how to build one that a Series A investor will trust.
Bottom-Up Revenue Model Structure:
Starting Point — Current ARR: Begin with your ending ARR today, not a target. Every line in the model flows from known inputs.
Component 1 — New Logo Revenue
Inputs:
- Marketing-sourced leads per month: [X]
- SDR-sourced leads per month: [Y]
- Lead-to-opportunity conversion: [%]
- Opportunity-to-close rate: [%]
- Average sales cycle: [months]
- ACV: [$]
Output: New ARR per quarter by segment
Component 2 — Expansion Revenue
Inputs:
- Current customer base: [N accounts]
- Net Revenue Retention rate: [%]
- Expansion triggers: seat growth, usage-based upsell, tier upgrade
Output: Expansion ARR added per quarter
Component 3 — Churn
Inputs:
- Logo churn rate (% of accounts): [%]
- ARR-weighted churn rate: [%]
Output: ARR lost per quarter
The 40% Growth Sanity Check:
Current ARR: $X
+ New logo ARR (4 quarters sum): $A
+ Expansion ARR: $B
- Churned ARR: $C
= Ending ARR: Should = X × 1.40
Where Models Break:
- Sales capacity: Can your current team generate the required pipeline? Model headcount as an input.
- Sales cycle timing: Revenue in Q4 requires opportunities created in Q2. Are they in pipeline today?
- NRR assumption: Every 5% change in NRR = ~$50K ARR swing at $1M base
Work backwards to validate: Target ARR × 1.40 - NRR-driven expansion = new logo target. New logo target / ACV = deals needed. Deals needed / win rate = pipeline needed. Are you generating that pipeline today?
What's your current ACV and win rate?
Get Started
How to install and use this skill in your preferred environment.
Skills are designed for AI coding agents (Claude Code, Cursor, Windsurf) and IDE-based workflows where the agent can read files, run scripts, and act on your codebase. Web-based AI can use the knowledge and frameworks, but won't have tool access.
Models & Context
Which AI models and context windows work best with this skill.
Recommended Models
Larger models produce more detailed, production-ready outputs.
Context Window
This skill's SKILL.md is typically 3–10 KB — fits in any modern context window.
All current frontier models (Claude, GPT, Gemini) support 100K+ context. Use the full window for complex multi-service work.
Pro tips for best results
Be specific
Include numbers — users, budget, RPS — so the skill can size the architecture.
Share constraints
Compliance needs, team size, and existing stack all improve the output.
Iterate
Start with a high-level design, then ask follow-ups for IaC, cost analysis, or security review.
Combine skills
Pair with companion skills below for end-to-end coverage.
Good to Know
Advanced guide and reference material for Financial Analyst. Background, edge cases, and patterns worth understanding.
Contents
SaaS Unit Economics Reference
| Metric | Formula | Healthy Benchmark |
|---|---|---|
| CAC | Total Sales & Marketing Spend ÷ New Customers Acquired (same period) | Varies by segment; track trend, not absolute |
| LTV | ARPU × Gross Margin % ÷ Monthly Churn Rate | 3–5x CAC minimum |
| LTV:CAC Ratio | LTV ÷ CAC | >3x; >5x is strong |
| Payback Period | CAC ÷ (ARPU × Gross Margin %) | <12 months (SMB), <18 months (mid-market), <24 months (enterprise) |
| Magic Number | Net New ARR ÷ Prior Quarter S&M Spend | >0.75 efficient; >1.0 very efficient |
| Rule of 40 | Revenue Growth Rate % + EBITDA Margin % | ≥40 for healthy SaaS |
Note on LTV: Use gross margin in the LTV calculation, not revenue. A 70% gross margin business with $1,000 ARPU has an effective LTV of $700/month compounding — not $1,000.
ARR Stage Benchmarks
| ARR Stage | Expected YoY Growth | Burn Multiple | Net Revenue Retention |
|---|---|---|---|
| $1M | 150–300% | <2x | >100% |
| $5M | 80–150% | <1.5x | >105% |
| $20M | 50–100% | <1x | >110% |
| $50M | 30–60% | <0.75x | >115% |
Burn Multiple = Net Cash Burned ÷ Net New ARR. Below 1x means you're generating more ARR than you're burning cash to create it.
Bottom-Up vs Top-Down Modeling
Top-down: Start from TAM, apply market share assumptions. Fast to build, but investors discount it heavily because the assumptions are unfalsifiable. Use it only for framing the opportunity size.
Bottom-up: Start from known inputs — sales rep count × quota attainment, or current pipeline × historical conversion × ACV. Every line is traceable to an operational assumption you can defend.
When to use each:
- Top-down: Investor pitch narrative, market sizing slide
- Bottom-up: Board financial package, fundraise model, annual planning
Building a credible bottom-up: For a product-led motion, use active users × conversion rate × ARPU × seats. For a sales-led motion, use qualified pipeline × win rate × average ACV ÷ average sales cycle length. Never mix the two approaches in the same forecast without clearly segmenting them.
DCF Assumptions for SaaS
Early-stage SaaS DCF analyses are mostly academic exercises — the terminal value dominates and is highly sensitive to assumptions. That said, understanding the conventions matters for investor conversations.
| Stage | Typical Discount Rate | Why |
|---|---|---|
| Seed | 40–60% | High execution risk, no proof of repeatability |
| Series A | 25–35% | Product-market fit established, GTM unproven |
| Series B+ | 15–25% | Repeatable growth, closer to comps-based valuation |
Terminal growth rate convention: Use 2–4% for mature SaaS (approximates long-run GDP growth). Using 5%+ will be challenged.
Why DCF matters less at early stages: At seed and Series A, the discount rate swing from 40% to 50% changes valuation by 30–40%. Investors at these stages rely more on comparable transaction multiples (ARR multiple) than DCF.
Common Financial Modeling Errors
| Error | Why It Matters |
|---|---|
| Hockey-stick without drivers | A revenue inflection in month 18 with no corresponding headcount, marketing spend, or product change is not a model — it is a wish |
| Ignoring churn in revenue projections | A 10% annual churn rate cuts your 5-year revenue projection by ~40% vs a zero-churn assumption |
| Mixing bookings and revenue | Bookings are signed contracts; revenue is recognized as earned. Swapping them overstates current-period revenue and violates GAAP |
| Missing payroll taxes in headcount model | Payroll taxes add ~15–20% to fully-loaded salary cost. Forgetting them understates headcount expense materially |
| Single-scenario forecast | Presenting one forecast number without a bear/base/bull case signals the model is a target, not an analysis |
Ready to try Financial Analyst?
Install the skill and start getting expert-level guidance in your workflow — any agent, any IDE.
$skillfish add ~skills/financial-analyst