Resources / PROFIT · Prompt
Profit simulator from your invoices
Claude turns your invoices into a live simulator. Raise a price, drop a client, kill a subscription — and watch profit move. Then it tells you which clients to fire. It's usually not who you'd think.
Setup — 2 minutes
- Export your invoices from your accounting tool as CSV or Excel (last 6–12 months). Columns you want: client, date, amount, paid date.
- Open Claude. New chat. Upload the file (or paste the rows).
- Paste the prompt and fill in the [brackets] — fixed costs and rough hours per client are what make it accurate.
- Hit go.
What you get
- A ranked list of clients by profit per hour — the one number most owners have never seen.
- Sliders that show what happens before you do it.
- Five moves, in order, with the euro effect of each.
Make it accurate
- Hours are the input that changes everything. Even a rough guess beats leaving it blank.
- Include the subscriptions you forgot about. Ask Claude: "Which of these costs would you cancel first and why?"
- Re-run quarterly. Ask it to compare with the last simulator.
Privacy note: you're uploading financial data. Use a plan where your data isn't used for training, remove client names if you prefer ("Client A, B, C"), and delete the chat after. The simulator works just as well anonymised.
You are my finance analyst. Build me a profit simulator from my real numbers. No motivational talk. DATA Invoices for the last [6–12] months: [attached file / pasted rows — columns: client, date, amount, paid date] Monthly fixed costs: [rent, salaries/contractors, tools and subscriptions — name + amount each] Rough hours per client per month: [Client A: 12h, Client B: 30h, ...] My target hourly rate: [€ ] Currency: [€] TASK 1. Clean the data. Group revenue by client and by month. Show me a short table so I can check you read it right. 2. For every client: revenue, hours, effective hourly rate, share of total revenue, average payment delay (if paid dates exist), trend over the period (growing / flat / shrinking). 3. Rank clients from most to least profitable PER HOUR. Flag: - anyone below my target rate - anyone above 30% of total revenue (dependency risk) - anyone who pays more than 30 days late - anyone shrinking for 3+ months 4. List every fixed cost and its share of monthly profit. Flag subscriptions that look unused or duplicated. 5. Build an INTERACTIVE SIMULATOR (HTML artifact) with controls for: - raise all prices by X% / raise one client by X% - drop client Y - cut cost Z - add or remove N hours on client W Show monthly revenue, costs, profit, margin and total hours updating live. Show the baseline next to the scenario. 6. Finish with THE FIVE MOVES you would make, in order, each with its profit effect per month and the risk of doing it. RULES - Use only my numbers. Where you must assume, show the assumption on the dashboard in plain sight. - If a client's hours are missing, mark their hourly rate "unknown" — do not invent. - Keep the five moves brutally specific: names, amounts, dates.
System.