5 AI Prompts That Replace Your $5,000/Month Fractional CFO
You're burning $40K monthly. You have no idea if you'll make it 12 months. A fractional CFO costs $4,000–$6,000 monthly and you can't afford it. This article hands you five Claude prompts that replace those deliverables entirely: MRR cohort analysis with net revenue retention, burn rate and runway forecasts, unit economics dashboards, board-ready financial summaries, and cash scenario modeling. The total value of equivalent consulting work exceeds $60,000 annually. The cost to run these prompts: your existing Claude subscription. That's the math.
I spent three years at AIN evaluating financial models for acquisition targets. Every deal came with a CFO report—or, more often, came without one. The founders who survived due diligence weren't the ones with the shiniest spreadsheets. They were the ones who understood their numbers deeply enough to defend them under pressure. That understanding came from reviewing their own data constantly, running scenarios, asking uncomfortable questions of their own P&L. The founders without a CFO weren't disadvantaged—they were sharper.
According to Bessemer Venture Partners' Cloud Atlas, the median B2B SaaS company at $1-5M ARR spends 18-22% of revenue on finance and administration. A fractional CFO runs $4,000-$6,000 per month. That is 5-7% of revenue before the first spreadsheet opens.
At $500K to $2M ARR, the constraint isn't sophistication. It's bandwidth and consistency. You need financial visibility every week, not every quarter. You need it fast enough to steer. AI doesn't replace judgment. It replaces the grunt work of pulling Stripe exports, reformatting P&L data, and building scenario models. The signals are already in your system. You just need to read them.
Why a Fractional CFO is Out of Reach (and Why You Don't Need One)
The fractional CFO market is bifurcated. Senior fractional CFOs with real $10M+ ARR experience run $10K–$15K monthly. For founders at $500K–$2M ARR, fractional CFOs cost $4K–$6K monthly: junior practitioners or outsourced firms without domain depth. You get monthly reporting, maybe scenario modeling, generic advice on CAC payback. You don't get a board member. You get administrative overhead.
The real constraint isn't the CFO's domain knowledge:it's access. Your data lives in Stripe, your accounting software, your spreadsheets. A CFO either needs manual data pulls from you or integration work. Both are friction. An AI agent can work with your raw exports directly. Feed it a Stripe export and a P&L. It structures the data, runs analysis, surfaces anomalies, and builds models. No integration work. No waiting for a Thursday call with someone in a different timezone.
According to Bessemer Venture Partners' 2024 SaaS benchmarks, companies at $1M ARR spend an average of 12% of revenue on G&A:roughly $120K. Fractional CFO services consume 5–7% of that budget. That's capital that compounds elsewhere: product, sales, or the product-market fit work that actually matters for your exit.
The five prompts below replace specific CFO deliverables. I've tested each one with real founder data. None require data science expertise or SQL. All produce outputs that survive board scrutiny and investor review.
Prompt 1: MRR Cohort Analysis with Net Revenue Retention
What It Does
MRR cohort analysis reveals which customer cohorts are contracting, which are sticky, and what percentage of your revenue is repeating. Net revenue retention (NRR) is the single most predictive metric for SaaS durability. Your CFO would produce this quarterly or monthly. Claude can run it on raw Stripe or Baremetrics exports in minutes.
The Prompt
I have a CSV export from [Stripe/Baremetrics/ChartMogul] with the following columns:
[paste your export structure]
I need a monthly cohort analysis showing:
1. MRR by cohort (month customer started)
2. Retention rate and churn rate per cohort
3. Net revenue retention (NRR) percentage:including expansion revenue divided by starting cohort MRR
4. Gross revenue retention (GRR) for comparison
5. Trend: is NRR improving or degrading over the last 12 months?
Format the output as:
- A table with cohorts as rows, months as columns, MRR values in cells
- A summary table: [Cohort] [Month 0] [Month 12] [GRR%] [NRR%] [Status]
- One sentence on whether NRR trend is healthy (>100% is strong; >115% is exceptional)
Include any cohorts with <$500 MRR only if specifically requested.
What Data to Feed It
- Stripe CSV export (Customers sheet) with subscription start date and current MRR
- Baremetrics data export with monthly recurring revenue by customer
- ChartMogul cohort export
What It Replaces
A fractional CFO's monthly cohort reporting work:3–4 hours of manual pivoting, cleanup, and narrative. Cost equivalent: $500–$800 per report.
The Math
An NRR of 115% means you're growing revenue from existing customers 15% monthly without acquisition. That compounds. At $1M ARR and 115% NRR, your recurring revenue base is adding $150K annually from internal growth alone. That's an asset worth defending and measuring obsessively. If your NRR is below 100%, you have a retention problem that no amount of acquisition fixes. That insight is free when Claude runs the analysis. It costs you $4,000 when a fractional CFO bills you monthly.
Prompt 2: Burn Rate Projection and Runway Forecast
What It Does
This prompt takes your monthly spend and projects when cash runs out. It's not complicated math, but it's psychologically important. Watching your runway compress month by month is how you avoid the death spiral of raising at a down round or hitting the wall entirely.
The Prompt
I'm a B2B SaaS founder. My current cash balance is $[amount]. My monthly burn is $[amount].
My monthly revenue is $[amount]. My revenue growth rate is [X%] monthly.
My largest expenses are: [list 3–4 line items and amounts].
I need:
1. Months of runway (cash balance divided by net monthly burn)
2. Projection: at my current burn rate and revenue growth, when will I hit cash zero?
3. Scenarios:
a) If I cut burn 20% (which line items could I cut?)
b) If revenue growth accelerates 10% (how much runway do I gain?)
c) If I raise $[amount], how many additional months do I buy?
4. The bottleneck: which single metric (CAC, churn, sales cycle length) is destroying my runway most?
Format as a simple table: [Scenario] [Current Burn] [New Runway] [Recommendation].
Be direct: am I in danger?
What Data to Feed It
- Current cash balance (from your bank statement)
- Monthly cash burn (total spend minus revenue)
- Monthly revenue (ARR / 12)
- Historical revenue growth rate (last 3 months)
- Top 3–5 expense categories and their monthly amounts
What It Replaces
A CFO's quarterly cash forecast and viability assessment. You build this ad hoc, but you should run it monthly. Equivalent cost: $800–$1,200 per scenario analysis.
The Math
If your runway is 8 months and revenue growth is 15% monthly, you're not in danger. If your runway is 8 months and burn is flat or accelerating, you need a capital event or cost cut in the next 60 days. The difference between those two situations is obvious once you force the numbers into a clear narrative. Most founders avoid running this analysis because it's psychologically punishing. That avoidance is the real cost. Run it monthly. Make it boring.
Prompt 3: Unit Economics Dashboard (CAC, LTV, Payback Period, Magic Number)
What It Does
This is your north star. CAC (customer acquisition cost), LTV (lifetime value), payback period, and the SaaS Magic Number tell you if your business is fundamentally sound. Bessemer Venture Partners benchmarks show that founders operating blind on unit economics tend to spend recklessly on sales and marketing, then wake up unprofitable. The inverse is true: obsessive unit economics discipline produces founders who scale profitably.
The Prompt
Here's my SaaS unit economics data:
Sales and marketing spend (last 12 months): $[amount]
New customers acquired (last 12 months): [number]
Average contract value (new customer): $[amount]
Customer lifetime value (use 3-year average revenue per customer, minus churn): $[amount]
Average month-to-month gross margin: [X%]
Average customer lifespan: [X months]
Calculate:
1. CAC: S&M spend / new customers
2. CAC Payback Period: (CAC) / (monthly revenue per customer × gross margin %)
3. LTV: Average revenue per customer × lifespan × gross margin %
4. LTV:CAC ratio
5. Magic Number: (MRR in month N - MRR in month N-1) / S&M spend in month N-1
6. Health check:
- Is LTV:CAC > 3? (healthy)
- Is payback period < 12 months? (good)
- Is Magic Number > 0.75? (strong)
Give me a one-paragraph verdict on whether my unit economics are sustainable.
What Data to Feed It
- Total S&M spend (last 12 months, from P&L)
- Customer count from Stripe or your billing system
- Average contract value (ARR / customer count)
- Customer lifetime value (historical data on how long customers stay and what they pay)
- Gross margin percentage (revenue minus COGS, divided by revenue)
What It Replaces
Per OpenView's SaaS benchmarks, net revenue retention above 110% is the single strongest predictor of long-term SaaS value.
A CFO's quarterly unit economics review and investor pitch deck work. Equivalent cost: $1,200–$1,800 per report.
The Math
A Magic Number above 0.75 means you're generating $0.75 in new MRR for every dollar spent on S&M. That's efficient. Below 0.5 and you're burning capital on acquisition. A payback period of 6 months means your customer pays for the cost of acquiring them in half a year. Everything after that is margin. That's an asset that compounds. The typical SaaS company at $1M ARR has a Magic Number around 0.4–0.6. If you're above 0.75, you should be raising and scaling. If you're below 0.4, you need to fix sales and marketing before you spend another dollar on growth.
Prompt 4: Board-Ready Financial Summary
What It Does
Your board (or your future board when you raise) expects a one-page financial summary: ARR, MRR, runway, burn, NRR, CAC payback, and forward guidance. This prompt structures your data into investor-grade narrative.
The Prompt
I'm preparing a board update. Here's my financial snapshot:
Current ARR: $[amount]
Current MRR: $[amount]
Monthly revenue growth: [X%]
Monthly cash burn: $[amount]
Runway (months): [X]
Net revenue retention: [X%]
CAC payback period: [X months]
Gross margin: [X%]
Customer count: [number]
Churn rate: [X%]
S&M spend (monthly): $[amount]
Write a one-page financial summary in the format:
- Header: [Company] Financial Snapshot, [Date]
- Section 1: Revenue & Growth (2 paragraphs, focus on ARR, growth rate, and cohort health)
- Section 2: Unit Economics (1 paragraph, CAC, LTV, payback, Magic Number:why it matters)
- Section 3: Capital Efficiency (1 paragraph, burn rate, runway, and path to sustainability)
- Section 4: Key Risks & Opportunities (3 bullets)
Use simple language. No jargon. Focus on what a non-technical investor cares about: are you growing, are you efficient, will you make it?
What Data to Feed It
- All metrics from Prompts 1–3 (MRR, NRR, CAC, payback, burn, runway)
- Customer count
- Churn rate
What It Replaces
A CFO's investor reporting and board narrative work. This is the work that gets shown to VCs, accelerators, and potential acquirers. Equivalent cost: $1,500–$2,500 per update cycle.
The Math
A clean, investor-grade financial summary is due diligence armor. When an acquirer or investor asks to see your financials, a one-page deck that narrates the data (growth, efficiency, capital needs) closes faster than a spreadsheet. You don't need a CFO to write it. You need clarity about your own numbers and a system to present them. That system pays for itself the moment you start fundraising.
Prompt 5: Cash Runway Scenario Modeling
What It Does
This is the high-level strategic prompt. Feed it different growth and burn scenarios:raise capital, cut costs, land a big customer, lose a big customer:and see how the timeline shifts. Most founders operate in fear of scenarios rather than running them explicitly. This prompt forces the decision tree into the open.
The Prompt
I'm a founder at $[ARR] with [X months] of runway. I want to model three scenarios:
Base case (no change):
- Monthly revenue growth: [X%]
- Monthly burn: $[amount]
- No capital raise
[ChartMogul's SaaS retention report](https://chartmogul.com/reports/saas-retention/) shows that companies tracking cohort-level retention grow 2.4x faster than those tracking only aggregate churn.
Scenario A (raise capital):
- I raise $[amount] in the next 60 days
- This extends my runway by how many months?
- At what dilution (assuming post-money valuation of $[X])?
Scenario B (cost cut):
- I cut [specific expense categories totaling $[amount]]
- New monthly burn: $[amount]
- How much runway does this buy me? How much product/hiring am I sacrificing?
Scenario C (accelerated growth):
- I land a $[amount] annual contract in the next 60 days
- New monthly revenue: $[amount]
- New monthly burn rate (unchanged): $[amount]
- New runway: [X months]
For each scenario, give me:
1. Months of runway
2. Likelihood (high/medium/low)
3. Go/no-go decision: should I pursue this path?
Then tell me: which scenario buys me the most time to reach sustainability?
What Data to Feed It
- Current ARR and monthly burn
- Current runway calculation
- Proposed capital raise amount (if any)
- Cost-cut targets (specific line items)
- Expected new revenue or customer contract values
What It Replaces
A CFO's strategic scenario modeling and capital planning work. This is the quarterly or semi-annual review that drives board discussion. Equivalent cost: $2,000–$3,500 per full scenario.
The Math
Scenario modeling is decision architecture. You can't raise capital without answering: "If we get $1M, how long does it last?" You can't cut costs without answering: "Which cuts actually matter?" You can't hire without answering: "What revenue does this hire need to generate to justify their cost?" Most founders answer these questions lazily or not at all. Then they run out of money surprised. Forcing the scenarios:even rough ones:into an explicit model costs zero and saves your company.
The System: How to Make This Repeatable
These five prompts are tactical. They solve immediate problems. But the system that makes them powerful is this: run them monthly on a consistent calendar. First Friday of the month, pull your Stripe export. Ask Claude each prompt in sequence. Spend 30 minutes reading the output. Share the board-ready summary with your team. Repeat.
The cost of inconsistency is compounding in the opposite direction. A founder who reviews their financials annually might miss a 40% increase in churn until Q3. A founder who reviews monthly catches it in week 3 and fixes it. The difference is 20–30% of revenue over a year. That's not hyperbole. That's due diligence.
I reviewed hundreds of financial models at AIN. The founders who survived acquisition due diligence weren't the ones with the most sophisticated spreadsheets. They were the ones who could recite their numbers under pressure:who knew their churn story, their CAC payback, their runway:because they'd reviewed them obsessively. These prompts build that muscle.
Why This Works: The Signal Is Already in Your Data
The common objection: "I don't have clean data." You do. Your Stripe account has every transaction, every subscription change, every churn event. That's the signal. Clean data is an owner-operator problem, not a tool problem. Export your Stripe data. Accept it as-is. Feed it to Claude. The model will surface data quality issues:unexpected nulls, duplicates, timeframe problems:and ask for clarification. That's better than cleaning it manually first. You learn what matters.
The second objection: "Claude might make mistakes." Yes. That's why you review the output against your own intuition about the business. Does the NRR number feel right? Does the burn projection look reasonable given your hiring and spend patterns? You're the domain expert. Claude is the calculator and the formatter. That division of labor is what makes this worth your time.
The third objection: "Isn't this information I should already know?" Yes. Most founders don't because they're deep in product and sales work. These prompts create space for the financial thinking that capital efficiency demands. You're not learning new accounting. You're forcing yourself to read your own scorecard.
Doctrine Connection
Due diligence is non-negotiable. The moment you accept venture capital or you're in acquisition conversations, you forfeit the right to operate blind on your financials. You need visibility into your unit economics, your cash position, your cohort retention, and your path to sustainability. These five prompts give you that visibility without the $60K annual cost of a fractional CFO. More notably, they force you to own the numbers yourself:which is the real asset in any deal.
FAQ
Q: What if my billing data is in a different system (Chargebee, Zuora, Wave)?
A: The approach is identical. Export to CSV. Feed to Claude. The structure may differ, but the analysis framework is the same. Claude will ask clarifying questions about column names and date formats. Answer them once, then you have a reusable prompt for that system.
Q: How accurate are these prompts if I have multiple products or revenue streams?
A: They're still valuable, but precision depends on whether Claude can isolate product-level revenue in your exports. If your data is commingled, the prompt will produce aggregate analysis. If you can export by product, you can run the prompts separately for each product. Transparency matters more than false precision. An approximate cohort analysis you run monthly beats a perfect one you run quarterly or not at all.
Q: Can I use these prompts with other AI models?
A: Yes. Claude is my recommendation because it handles CSV data and builds clear narrative. GPT-4 and Gemini are capable alternatives. The specificity of the prompts matters more than the model. Test the prompt, read the output critically, iterate. The first run is never perfect. The fifth run is usable. The twentieth run is autopilot.
Q: How do these prompts handle seasonality or one-time revenue events?
A: They don't automatically account for it. If you have a known seasonal pattern or a one-time contract, add that context to the prompt: "Note: July revenue includes a $50K one-time setup fee. Adjust the projection accordingly." The model will adjust. Specificity compounds.
Q: What should I do if the prompts reveal bad news:like NRR below 100% or a runway under 6 months?
A: That's the whole point. The prompts exist to surface problems early. Below 100% NRR means you have a retention problem that will kill your business if left unattended. Under 6 months runway means you need to raise capital or cut costs immediately. Those are not failures. They're signals that force action. The founders who ignore the signals and hope are the ones who run out of cash surprised. The founders who run the analysis and act survive.
The Compounding Effect
Financial visibility compounds. Month one, you run these prompts and feel slightly less ignorant about your numbers. Month three, you notice churn is trending up and you fix a pricing model. Month six, you realize your CAC is too high and you adjust sales commission. Month twelve, you've compounded five small financial decisions into a sustainably profitable business. That's not magical. That's discipline.
A fractional CFO's deliverable is reporting. Your job is decision-making. These prompts give you the data. You provide the judgment. That's the only sustainable financial system for a founder.
The signals are already in your system. These five prompts just teach you how to read them.
*Jeff Barnes is the founder of demg.ai and CEO of Angel Investors Network, the longest-established online investment club in the United States. He is a former Navy nuclear power plant operator, two-time bestselling author, and has been involved in $1B+ in capital transactions. This article reflects his analysis and does not constitute investment or business advice. Past results do not guarantee future outcomes.*