Home · Operations · AI automations
AI for business reports and financial analysis: 10 real small-business examples
Updated October 18, 2026 · 10 sourced examples · no paid placement
Short answer: Small business owners take reports they already have (a POS order export, monthly P&Ls, payroll summaries, salon booking reports, sales analytics) and have ChatGPT, Claude or Gemini turn them into a weekly report, a trend review or a budget. In most of these 10 examples it is a manual upload with a saved prompt, not a software project; one law firm connected Claude to QuickBooks. The owners treat the output as a second opinion: a brewery co-founder says every recommendation still needs human judgment. The results are what each business reported, not promises.
Below are 10 documented examples from a pizza restaurant group, a brewery, a seafood restaurant, a 151-year-old apothecary, a bath and body boutique, a nail salon, an optometry practice, a landscaping company, a law firm and a realtor, told by owners in their own words, in testimony to Congress, on podcasts and in the trade and business press: who did it, what the AI reports on, what they reported and where the story came from. After that: whether you can trust the analysis, what to watch out for and how to try it this week.
Other industries too: see real AI automations by task for any small business.
What these 10 businesses actually did
- A weekly ticket-time report from a POS export: Old Scratch Pizza (Ohio) (owner-told, via Toast Community: Exploring AI in My Restaurant, Ask Me Anything (Eric, owner of Old Scratch Pizza)). Each week the team exports Toast's Order Details report for multiple restaurants, a CSV of over 8,000 lines, and uploads it to ChatGPT with a fixed prompt. ChatGPT works out each check's time from the earliest Fired Date to the latest Fulfilled Date and reports the share of checks done in 20 minutes or less per location, with a bar graph by day of week. Fixed prompts are kept in a folder in ChatGPT, and the process is a manual download and upload. He also uploads his P&L for an analysis and recommendations. Reported: The report is part of the weekly manager review and gives teams a specific measurement to work toward each week. "I could have never done this without AI," he says. No before-and-after ticket times are given. Full case
- Monthly P&L trends and KPIs: Vice Beer (Washington) (press, via Brewer Magazine: Treat AI Like Your Brewery's Junior Analyst, Not Its CEO). He uploaded monthly profit-and-loss statements into ChatGPT and asked it to review three months of data for trends. He now often asks it to take on the role of his CFO and hold him accountable on spending and payroll, and has refined his prompts over several months to get consistent reports. He also feeds it wholesale invoices, production numbers and retail sales to track sales channels and monthly KPIs and to weigh production decisions, and used it to flag clauses and questions in a long operating agreement before meeting his attorney. Reported: He says he was "absolutely blown away" by the trends it spotted that he hadn't noticed. In one case it pointed out that seven barrels of a Saison would tie up 15 kegs for five to six months, an opportunity cost nobody had raised. It has fed into pricing discussions, sales goals and KPI tracking, but he says every recommendation still needs human judgment about the local market. He refuses to use it for recipes or major branding. No dollar figures are given. Full case
- A weekly payroll check against your own benchmarks: Cowbell Seafood & Oyster (Washington, D.C.) (press, via James Beard Foundation: 10 Tips on How to Use AI Tools in the Restaurant Business (JBF Impact Team)). Every week he uploads payroll summaries to an AI tool with clear instructions to check them against defined benchmarks, plus a checklist of specific items to verify, so abnormalities get flagged. During a lease negotiation he uses AI as a partner to compare the pitfalls it sees with the ones he sees. The recap doesn't name the AI tool. Reported: "I can compare the pitfalls it sees with the pitfalls I think I see. That can wind up saving you big time in the long term," he says. The James Beard Foundation recap suggests adding your tip model and other benchmarks to catch errors before they hit your bottom line. No savings figure is given. Full case
- CFO-style analysis of the financial statements: Merz Apothecary (Chicago) (owner-told, via Testimony of Anthony Qaiyum to the House Small Business Committee, 'AI on Main Street'). He securely uploads the company's financial statements to Claude or Gemini and asks for analysis of cash flow, inventory investment, operating expenses, payroll trends and potential financial risks. The team is also building an AI agent to handle routine accounts payable so that employee can work on strategic problems. Reported: Self-reported: he gets 'a thorough analysis' of the statements without a CFO on staff. The accounts payable agent was still being built at the time of the testimony, so no time savings are reported yet. Full case
- A Monday review with Claude connected to QuickBooks: Business Law Group (New Orleans) (owner-told, via Endeavors AI podcast: Above the AI Cloud, how a boutique New Orleans firm went AI-first (Amanda Butler Schley)). Every Monday she consults three AI 'agents' set up as a CMO, CEO and CFO, giving them the firm's real reports. She uploads transcripts of meetings with her human outsourced CFO so the AI's advice stays consistent with his. Claude is connected to QuickBooks for financial analysis. An AI receptionist is in testing. For content she now questions every sentence, after another lawyer flagged the invented law in a published post and she had to pull it. Reported: Self-reported, no numbers: she says connecting Claude to QuickBooks unlocked analysis the firm never had. Lessons she shares: AI can 'say the law wrong' even on simple summaries; models tend to adopt your stance and feed it back (contrarian prompts haven't fixed that for her); the same question asked ten times gives different answers. Her team is skeptical of the AI receptionist so far. Full case
- A yearly budget built with a ChatGPT project: Landscape Endeavors (Colorado) (press, via Lawn & Landscape: 2026 Tech Report (Cory Shorette, Landscape Endeavors)). A project in paid ChatGPT is told "You are my CFO", given what the business does and its parameters, and loaded with financials, spreadsheets and budgets. Its standing instructions make it pressure-test every answer and list the assumptions it still needs confirmed. Reported: Per Shorette, it helped him build the 2026 budget with detailed overhead line items the company can stick to, and it's set up to assist with M&A questions. He says owners usually know the right decision, "but it's nice to have a double-check and backup of your thesis." Managers also use ChatGPT for professional emails about difficult situations. No dollar figures are published. Full case
- Sales analytics and a monthly recap for a one-woman shop: Mainland Vintage (Brooklyn) (press, via Business Insider: How an artist built an AI C-suite to help run her business). Since February 2026 she has run separate Claude projects on the $20/month plan for marketing, sales, finance and product development, each told to defer to her as CEO. The 'CFO' project interprets sales analytics but makes no decisions, a marketing chat handles Instagram strategy, and an AI 'chief of staff' compiles a monthly recap across the departments. She doesn't use AI for visual design. Reported: Self-reported: 'Seventy-five percent of the stuff I'm doing now with Claude, I wouldn't have been able to do before.' She stays on the $20 plan so the usage limit keeps her mindful, and says she would rather hire part-timers than buy more AI. Full case
- Break-even booking goals from salon reports: Cape Cod Nail Co. (Massachusetts) (owner-told, via AI for Growth: Lessons from Leaders podcast (transcript)). She gives ChatGPT her salon reports (booked hours, cost per hour and revenue per hour by season, recurring clients, retention) and works backwards in a layered back-and-forth, challenging every number, then has it produce a visual PDF of the goals. Reported: Self-reported: the analysis showed client counts held up in winter but higher-value pedicures fell off, which led to a winter pedicure marketing campaign. No revenue figures given. Full case
- True profit per product after rebates: Elite Eye Care (North Carolina) (owner-told, via Review of Optometric Business: 5 Data Points to Input Into ChatGPT for Contact Lens Profitability (Haley Perry, OD)). On ChatGPT 4.0 ($20/month at the time she wrote), she uploads cost fees, MSRP or office prices, sales volume and manufacturer rebate details, then asks for savings if she shifts market share, true cost per box after rebates, the most profitable lenses, and expected rebate checks. She also feeds industry and competitor pricing to set prices, photographs trial inventory for a restock list, and uses purchase-pattern analysis for promotional content. She organizes and cleans the data before asking for trends. Reported: Self-reported, no dollar figures published: she says ChatGPT 4.0 became 'an integral part of our strategy to enhance the profitability of our contact lens segment' and that image upload made trial inventory faster than a scanner. Treat rebate and ROI math as a first pass and check it in a spreadsheet. Full case
- Neighborhood market reports for clients: Missy Plunkett, realtor (South Carolina and Tennessee) (press, via Accessible AI newsletter). ChatGPT drafts listing descriptions, personalised follow-ups using past conversations with each client, and neighbourhood market reports from data she uploads. Reported: Self-reported: a listing description takes about 20 minutes instead of hours, about 40 hours a month saved overall, and client response rates up almost 30%. Full case
These are each company's own results with their own tools and teams, as reported by the source. Vendor case studies are written by the company selling the tool. None of them are promises.
What reports can AI build from data you already have?
Weekly operating reports. Each week the team at Old Scratch Pizza in Dayton, Ohio, exports Toast's Order Details report for its restaurants (a CSV of over 8,000 lines) and uploads it to ChatGPT with a fixed prompt. ChatGPT works out each check's time and reports the share done in 20 minutes or less per location, with a bar graph by day of week, and the report is part of the weekly manager review. The owner, Eric, says Toast gave him only rudimentary fulfillment data: "I could have never done this without AI." Reid Schilling of Cowbell Seafood & Oyster in Washington, D.C., uploads payroll summaries every week with instructions to check them against his own benchmarks, plus a checklist of items to verify, so abnormalities get flagged.
Monthly financial reviews. Michael Perozzo, co-founder of Vice Beer, a brewery in Vancouver, Washington, uploaded monthly profit-and-loss statements to ChatGPT and asked it to review three months for trends. He now has it play CFO on spending and payroll and track sales channels and monthly KPIs; in one case it pointed out that seven barrels of a Saison would tie up 15 kegs for five to six months. Anthony Qaiyum, who runs Merz Apothecary in Chicago, uploads the store's financial statements to Claude or Gemini for an analysis of cash flow, inventory investment, operating expenses, payroll trends and risks, because the business can't afford a full-time CFO.
A standing 'CFO' to check in with. Amanda Butler Schley of Business Law Group in New Orleans consults AI set up as a CMO, CEO and CFO every Monday, using the firm's real reports and transcripts of meetings with her human outsourced CFO. Claude is connected to QuickBooks, which she says unlocked analysis the firm never had. Cory Shorette of Landscape Endeavors in Colorado Springs set up a paid ChatGPT project told "You are my CFO", loaded with financials and budgets and instructed to pressure-test every answer; it helped build the company's 2026 budget with detailed overhead line items. Sara Russell runs a Claude 'CFO' project that interprets sales analytics for Mainland Vintage, her Brooklyn bath and body boutique, and an AI 'chief of staff' that compiles a monthly recap.
Pricing and booking goals. Sarah Mason, who owns two Cape Cod Nail Co. salons, gives ChatGPT her salon reports (booked hours, cost and revenue per hour by season, retention) and works backwards to break-even booking goals. It showed client counts held up in winter but higher-value pedicures fell off, which led to a winter pedicure campaign. Dr. Haley Perry of Elite Eye Care in Arden, North Carolina, uploads contact-lens costs, sales volume and rebate details to ChatGPT to see the true cost per box and which lenses are most profitable.
Reports for clients. Missy Plunkett, a realtor in South Carolina and Tennessee, has ChatGPT draft neighborhood market reports from data she uploads, along with listing copy and client follow-ups, and reports about 40 hours a month saved overall. At The Advisers, an insurance brokerage in New Zealand, the firm's platform pulls a client's last 12 months of notes and correspondence into one summary before a renewal conversation, which director Nic Tyson says can cut 30-plus minutes of preparation in half.
Can you trust the analysis?
Treat it as a second opinion. Perozzo says every recommendation still needs human judgment about the local market. Shorette says owners usually know the right decision, "but it's nice to have a double-check and backup of your thesis." Mason challenges every number in a back-and-forth before she accepts the goals.
It can be wrong, and it can agree with you too easily. Schley learned this the hard way: a ChatGPT summary of a legal education session invented the law, the post went live, and she had to pull it. She also says models tend to adopt your stance and feed it back, and the same question asked ten times gives different answers. Check any math, such as rebate or payback figures, in a spreadsheet before you act on it.
What to watch out for
Clean, consistent data first. Perry organizes and cleans her data before asking for trends, and Tyson says consolidating the firm's data into "one source of truth" came first. The Old Scratch report works week after week because the same export goes in with the same saved prompt, and Perozzo refined his prompts over several months to get consistent reports.
What you upload. Qaiyum says he uploads statements securely. Leave out names and account numbers the analysis doesn't need, and check your AI tool's data settings before you upload financials.
Tools named in these examples
| Tool | Starting price (base plan) | Checked |
|---|---|---|
| ChatGPT | Free $0; Go $8/mo; Plus $20/mo; Pro $100/mo | 2026-10-17 |
| Claude | $17/mo billed annually ($20 monthly) | 2026-10-15 |
| Google Gemini | $4.99/mo (Google AI Plus); free plan available | 2026-10-15 |
| QuickBooks | $38/mo (Simple Start; free plan and intro discounts available) | 2026-10-17 |
| Toast | POS from $0/mo (Starter Kit) or $69/mo; plus $9/employee/mo on some plans | 2026-10-14 |
Starting prices from each vendor's pricing page on the date shown. AI phone and texting features can be priced separately from the base plan, so check before you buy. Mangomint and JAVLN Officetech are also named in these examples; we haven't price-checked them.
How to start this week
- Pick one question you ask every week or month: which location is slowest, where payroll is drifting, which products actually make money.
- Find the export that answers it: a POS order report, your P&L, a payroll summary, your booking software's reports.
- Write one prompt that says exactly what to calculate and what to show (a table, a chart, the three things to look at), and save it so every run is the same.
- Upload the export with the prompt, then check a few numbers by hand before you share the report. Leave out names and account numbers it doesn't need.
- Run it on the same day each week or month and bring it to your manager meeting. Once it is routine, an automation can pull the export for you.
Upload a Toast order export to ChatGPT for a weekly ticket-time report across locations
Each week the team exports Toast's Order Details report for multiple restaurants, a CSV of over 8,000 lines, and uploads it to ChatGPT wi…
Use ChatGPT as a brewery's junior analyst: upload monthly P&Ls and sales data and ask it to play CFO
He uploaded monthly profit-and-loss statements into ChatGPT and asked it to review three months of data for trends.
Check weekly payroll against your own benchmarks with an AI checklist, and pressure-test a lease
Every week he uploads payroll summaries to an AI tool with clear instructions to check them against defined benchmarks, plus a checklist …
Get CFO-style analysis of your financial statements from AI and build an agent for routine accounts payable
He securely uploads the company's financial statements to Claude or Gemini and asks for analysis of cash flow, inventory investment, oper…
Run a weekly review with an AI 'C-suite' (CMO, CEO, CFO) fed your real reports, with Claude connected to QuickBooks
Every Monday she consults three AI 'agents' set up as a CMO, CEO and CFO, giving them the firm's real reports.
Set up a ChatGPT project as a virtual CFO that builds the yearly budget and pressure-tests its own assumptions
A project in paid ChatGPT is told "You are my CFO", given what the business does and its parameters, and loaded with financials, spreadsh…
Run a one-woman shop with a Claude project for each department and a monthly AI chief-of-staff recap
Since February 2026 she has run separate Claude projects on the $20/month plan for marketing, sales, finance and product development, eac…
Find the gaps in your salon schedule and set booking goals with ChatGPT
She gives ChatGPT her salon reports (booked hours, cost per hour and revenue per hour by season, recurring clients, retention) and works …
Upload contact-lens costs, rebates and sales volume to ChatGPT so you can see true profit per box and set prices
On ChatGPT 4.0 ($20/month at the time she wrote), she uploads cost fees, MSRP or office prices, sales volume and manufacturer rebate deta…
Use ChatGPT for listing copy, client follow-ups and market reports
ChatGPT drafts listing descriptions, personalised follow-ups using past conversations with each client, and neighbourhood market reports …
Questions
Can AI write business reports from my data?
Yes, from exports you already have. Old Scratch Pizza turns a weekly Toast order export into a ticket-time report by location, Cowbell Seafood & Oyster checks weekly payroll against its own benchmarks, and Mainland Vintage gets a monthly recap across marketing, sales, finance and product. A person still reads and checks each one.
Can I use ChatGPT for financial analysis?
Owners here do. Vice Beer reviews monthly P&Ls with ChatGPT, Landscape Endeavors built its 2026 budget with a ChatGPT project set up as a CFO, and Elite Eye Care works out contact-lens profit per box. Merz Apothecary uses Claude or Gemini, and Business Law Group connected Claude to QuickBooks. They treat it as a second opinion, not the final word.
Can AI analyze my financial statements?
It can give you a first read. Anthony Qaiyum says he gets "a thorough analysis" of Merz Apothecary's statements without a CFO on staff, and Michael Perozzo says ChatGPT spotted trends in Vice Beer's P&Ls that he hadn't noticed. Both are self-reported, and Perozzo says every recommendation still needs human judgment.
Do I need special software for this?
Mostly not. In most of these examples it is a manual export and upload with a saved prompt. Business Law Group connected Claude to QuickBooks, and The Advisers uses reporting built into its insurance platform.
Can someone set it up for us?
Yes. The $197 Quick-Win sets up one automation in accounts you own, live within 48 hours of getting access, with a refund if it is not live in 48 hours or is not doing its job in the first 14 days.
Related guides
- AI automations for any small business, by task
- AI for bookkeeping and invoice processing: real examples
- AI automations for small accounting firms and bookkeepers
- AI for proposals, quotes and estimates: real examples
- AI email triage and inbox management: real examples
- AI for restaurants: operations
- AI bookkeeping and admin tools for restaurants
- AI for shops: bookkeeping and admin
- Have one automation set up for you ($197)
One short email a day: 3 new AI use cases from real small businesses, each linked to its source, plus a Tool of the Day. Free, no spam, unsubscribe anytime.
Want one set up for you?
The $197 Quick-Win: we set up one AI automation that saves you hours each week, in accounts you own, live within 48 hours of getting access. No call needed. Refund if it isn't live in 48 hours or isn't doing its job in the first 14 days.
See the $197 Quick-Win Bigger setups