How to use AI to prepare a simple business performance summary you can share with a business partner, investor, or advisory board

Most small business owners with partners or advisors send nothing — or send a chaotic email thread with three spreadsheets attached. This post walks you through a repeatable workflow that turns your raw numbers and notes into a clean, readable AI business performance report your partner can actually use. The workflow works because AI converts a table of figures into plain-English narrative — the format non-finance readers actually absorb.

What you need before you start

Tool name: ChatGPT{:target="_blank"} (GPT-4o model) — accepts pasted data, bullet notes, and CSV content; converts them into structured narrative. Free tier available; GPT-4o requires ChatGPT Plus at $20/month. If you work in Google Sheets, Google Gemini{:target="_blank"} integrates natively with Google Workspace and removes the copy-paste step. For longer data exports or multi-month comparisons, Claude{:target="_blank"} (3.7 Sonnet) handles up to 200,000 tokens in a single session — useful if you're feeding in a full year of data.

Time required: 60–90 minutes the first time; 60–75 minutes once the workflow is established

Skill level: No coding required. Basic comfort copying data from a spreadsheet or accounting export helps.

---

Pull your raw inputs in under 30 minutes

Your AI business performance report is only as useful as what you feed it. Gather these inputs before you open any AI tool.

1. Open your accounting software — QuickBooks{:target="_blank"}, Xero{:target="_blank"}, or a spreadsheet — and pull the current month's figures. You should see revenue, cost of goods or direct costs, and your cash balance.

2. Record these six numbers in a plain text file or notes app: revenue this month, revenue target, gross margin (or gross profit), cash on hand, accounts receivable outstanding, and any significant one-time items. You should have six labeled figures, nothing else.

3. Write 3–5 bullet points covering what happened this month. Include one key win, one key challenge, and 1–3 priorities for next period. Don't edit yet — raw notes work fine here. You should have a rough list that covers the operational picture the numbers don't show.

4. Export one CSV from your accounting software if you want the AI to cross-check your figures. A profit and loss summary export works. You should see a file you can open in a text editor or paste directly into the AI chat.

These inputs take 20–30 minutes for most owners. The accounting export is optional — pasted numbers work just as well.

---

How to write a business update for a partner using AI

This is where 10 minutes of prompting replaces two hours of writing.

5. Open ChatGPT (or Claude, or Gemini) and start a new conversation.

6. Paste your six figures and your bullet-point notes into the chat window first. Do not ask it to do anything yet. You should see your raw data sitting in the input field.

7. Type your prompt below the data. Use this exact structure:

You are helping me write a monthly business performance summary for my business partner. My partner is not an accountant — they have a general business background and want to understand how we're doing without wading through spreadsheets.

Using the data and notes I've pasted above, write a one-page business performance summary that includes:

• A 2–3 sentence executive overview of the month
• Revenue vs. target (explain the gap or beat in plain English)
• Gross margin note (flag if it moved more than 2 points)
• Cash position (brief, no jargon)
• Key wins this month (2–3 bullets)
• Key challenges this month (1–2 bullets)
• Priorities for next period (up to 3 bullets)

Keep the language plain. No accounting jargon. Use a narrative paragraph for the overview, then bullets for the rest. Keep the total length under 400 words.

After the summary, list 3 questions a skeptical advisor might ask after reading this.

8. Review the output. You should see a structured summary with an overview paragraph, labeled sections, and three advisor questions at the bottom.

9. Edit the output for accuracy. AI occasionally smooths over a number or misreads a relationship in the data. Read every figure against your source. You should confirm that every number in the summary matches your inputs exactly.

The advisor questions step isn't decoration. It surfaces gaps in your own explanation — and prepares you for the actual conversation.

---

A real before-and-after

Before: An email sent to a silent business partner in March 2024. Subject line: "March update." Body: a screenshot of a QuickBooks P&L, two sentences about a slow month, and a note that payroll was delayed.

After (AI-generated summary):

March came in at $41,200 against a $48,000 target — a 14% miss driven by two delayed client invoices that pushed into April. Gross margin held at 62%, consistent with February. Cash position is $18,400, which covers the next 45 days of operating costs at our current run rate.

Wins: Closed a new retainer client ($3,200/month starting May). Resolved the supplier dispute from Q1. Challenges: Two invoices delayed by client approval processes, not project delivery. Pipeline for Q2 is thin below $10K deals. Priorities: Follow up on April AR by the 10th. Present Q2 pipeline to partner by April 20. Evaluate whether to hire a part-time ops coordinator.

Same data. Same month. One version takes three minutes to read and understand. The other requires a conversation to decode. That's the difference the narrative format makes for non-finance readers — a point Harvard Business Review has documented{:target="_blank"} in how business narratives influence decision-making.

---

Build this into a repeatable monthly workflow

The full process fits into four blocks on one day each month:

Block	Task	Time
Pull numbers	Export from accounting software, record 6 figures	30 min
Draft notes	Write bullet-point wins, challenges, priorities	20 min
Prompt AI	Paste data + prompt, review output	10 min
Edit and send	Check figures, adjust tone, send to partner	10–20 min

Set a recurring calendar event on the 2nd or 3rd business day of each month. That's when your prior month data is clean enough to use. The Xero Small Business Insights research{:target="_blank"} found that 60% of small business owners cite financial reporting as a significant time drain. This workflow doesn't eliminate the task — it compresses it.

---

When something goes wrong

The summary includes a number that doesn't match your data. AI occasionally interpolates or rounds figures when the input data is ambiguous. The cause is usually unlabeled numbers — the AI guessed what the figure represented. Fix: label every number in your input ("Revenue this month: $41,200") before pasting.

The output sounds too formal or too casual for your partner. The prompt didn't specify tone. Add one sentence to your prompt: "Write in a direct, professional tone — not corporate, not casual." Run it again.

The AI flags no challenges and makes everything sound positive. Your bullet-point notes didn't include any challenges, or you framed them too softly. Go back to step 3, write one honest challenge sentence, and re-prompt. AI reflects the inputs you give it. (This is not a bug.)

---

What to do next

Send this month's update. Don't wait until it's perfect — a clean one-page summary sent this week builds more trust with your partner than a polished report sent next quarter. Once you've run the workflow twice, look at how to use AI to set up a simple financial tracking system so your input data is already organized when you sit down to write.

---

FAQ

Can I use this workflow to write an investor update for a funded startup? This workflow is built for small business partner updates — owners with a silent partner, a co-owner, or an informal advisory board. Funded startup investor updates (the kind Y Combinator{:target="_blank"} recommends) include MRR, churn, burn rate, and runway metrics that don't apply to most service businesses or retail operations under $2M revenue. Use this workflow for its intended audience.

Is it safe to paste my financial data into ChatGPT? Paste aggregate figures — monthly revenue totals, margin percentages, cash balances — not individual transaction records, customer names, or bank login details. Aggregate numbers are sufficient for generating a useful summary and carry no meaningful privacy risk. Avoid putting personally identifiable customer data or account credentials into any public AI tool.

What if my partner wants to see the actual spreadsheet, not just a narrative? Send both. Attach your P&L export as a supporting document and lead with the AI-generated narrative in the email body. Most non-finance partners read the narrative and reference the spreadsheet only if a number surprises them. The narrative is the summary; the spreadsheet is the backup.

How do I handle months where the business had a bad result? Write the challenge bullet honestly in step 3, then add this line to your prompt: "Don't soften negative results — explain them clearly and include what action we're taking." AI defaults to diplomatic framing unless you instruct otherwise. A clear explanation of a bad month — with a response plan — builds more partner trust than a vague one.

Do I need a different prompt for a quarterly or annual summary? Adjust two things: the time period referenced in the prompt, and the data you paste. For a quarterly summary, include three months of figures and note any trend direction. Add this to your prompt: "Identify any trend across the three months — improving, declining, or stable — and call it out in the overview paragraph."