Best Ways Use AI in Finance: 6 Proven Strategies Every Finance Professional Can Implement Today

• 📹 Watch the Complete Video Tutorial
• 1. Use Generative AI to Identify Your Most Valuable KPIs (Traditional + Innovative)
• The Exact Prompt to Use
• How to Feed Your Data to AI
• Why Use GPT-4 (or “o3”/“03”) Model?
• Example Output: Traditional vs. Innovative KPIs
• 2. Run Cohort Analysis in Seconds (Not Hours)
• What Is Cohort Analysis?
• How AI Automates It in Under a Minute
• No Data Upload? No Problem
• 3. Automate Data Cleaning with AI-Generated Google Apps Script
• Real-World Example: Credit Card Statement Consolidation
• Step-by-Step AI Automation Process
• Advanced Use Case from AI Finance Club
• 4. Ask AI: “What Don’t I Know That I Don’t Know?”
• How to Set Up a Custom AI Assistant
• Surprising Insights AI Can Reveal
• Real Example: New Hire in Entertainment Park Industry
• 5. Leverage AI-Native Finance Tools: Puzzle & Concourse
• Puzzle: AI-Powered Accounting for Startups
• Concourse: AI Reporting for QuickBooks Users
• Why AI-Native Tools Win
• 6. Use AI to Generate API Code for Automated Data Pulls
• Step-by-Step: Automate QuickBooks Management Reports
• Prompt Strategy: Ask for Options First
• AI’s Recommended Solutions
• Deep Dive: Generate the API Code
• Sample Code Snippet (as generated by AI)
• Choosing the Right AI Model Matters
• Security & Data Privacy Best Practices
• Real Results: Time Savings & ROI
• Tools Mentioned in This Guide
• Who Should Use These Strategies?
• Getting Started: Your First AI Finance Project
• Free Resource: AI for Finance Course
• Final Thoughts: AI Is Your Finance Co-Pilot
• Ready to Transform Your Finance Workflow?

You don’t need a data scientist on your team to start using AI in finance. In fact, hundreds of finance teams have already learned how to leverage AI to automate manual tasks, clean messy data, and even generate actionable insights in seconds. This comprehensive guide walks you through the six exact strategies used by finance professionals to integrate AI into their daily workflows—without writing a single line of code.

From running cohort analyses in under a minute instead of four hours, to automating data cleaning with Google Apps Script generated by AI, to uncovering blind-spot KPIs you never knew you needed—this article covers every technique, tool, example, and step-by-step process shared in the original transcript.

Let’s dive into the best ways use AI in finance—starting today.

1. Use Generative AI to Identify Your Most Valuable KPIs (Traditional + Innovative)

One of the most powerful ways to begin using AI is by asking it to help you identify the right metrics to track. Instead of guessing, use a structured prompt to extract both classic and forward-thinking KPIs tailored to your specific dataset.

The Exact Prompt to Use

Here’s the proven prompt used by finance professionals:

“What are the five traditional KPIs and five innovative KPIs to track my performance based on this data? Tell me why and how to calculate this in Excel and which graph could help me.”

How to Feed Your Data to AI

To make the AI understand your context:

1. Open your dataset (e.g., a supermarket sales table with columns like Product, Units Sold, Revenue, Margin, etc.)
2. Copy the first 5 rows of data (including headers)
3. Paste it directly into ChatGPT (or your preferred AI tool)

Pro Tip: If you’re using enterprise tools like ChatGPT Teams, Copilot with a professional license, or similar secure platforms, you may be able to upload your full confidential file directly—always verify your company’s data policy first.

Why Use GPT-4 (or “o3”/“03”) Model?

The speaker recommends switching to a more advanced model (referred to as “03” or likely GPT-4) because:

• It takes more time to think (up to 1 minute 18 seconds in the example)
• It provides deeper reasoning, visible via “Expand” to see the thought process
• It generates not just KPIs, but also Excel formulas and visualization recommendations

Example Output: Traditional vs. Innovative KPIs

For a retail dataset, AI generated the following:

This single prompt delivered 10 actionable KPIs with implementation guidance—something that could take hours to research manually.

2. Run Cohort Analysis in Seconds (Not Hours)

Cohort analysis tracks user behavior over time—critical for subscription businesses like Netflix. Traditionally, building this in Excel takes at least 4 hours due to complex calculations, data reshaping, and conditional formatting.

What Is Cohort Analysis?

Example: Group customers by their month of subscription, then track what % remain active after 1, 2, 3… months. A table might show:

• January 2022 cohort: 100% retained at Month 1, 94% at Month 5, only 42% at Month 23
• An outlier like August 2022 might show unusually high or low retention

This reveals whether your retention strategy is improving or degrading.

How AI Automates It in Under a Minute

Follow these steps:

1. Prepare your raw data: columns like Date, Customer ID, Product, Invoice
2. Upload the file directly to ChatGPT (if allowed by your security policy)
3. Use this prompt: “Can you do a cohort analysis visually, by months, on the retention rate?”

ChatGPT will:

• Read your data using Python
• Generate the full cohort table with retention percentages
• Output clean, formatted results—ready to visualize
• Show the Python code used, so you can audit the logic

Time saved: From 4 hours → 30–60 seconds.

No Data Upload? No Problem

If you can’t upload sensitive data:

• Copy the generated Python code
• Run it in your own secure environment (e.g., Jupyter Notebook, internal server)
• Input your local data file

3. Automate Data Cleaning with AI-Generated Google Apps Script

Finance teams often waste hours manually cleaning messy data—like credit card statements with multiple tabs, irrelevant headers, and inconsistent formatting. AI can now write the automation script for you.

Real-World Example: Credit Card Statement Consolidation

Problem: A Google Sheet contains multiple tabs—one per employee—with credit card transactions. Each tab has:

• Unneeded header rows
• Extra columns (only C, D, E are useful)
• No clear link to cardholder or company name (stored in cell B2 of each tab)

Step-by-Step AI Automation Process

1. Sample your data: Copy the first few rows from one tab and paste into ChatGPT
2. Describe your goal: Use this prompt:
   “I want to: (1) consolidate all tabs into one sheet, (2) keep only columns C, D, E, (3) delete messy header rows, (4) retain cardholder name and company name from cell B2 of each tab.”
3. Request the script: Add: “Write the Google Apps Script for me.”
4. Get the code: ChatGPT outputs a complete, ready-to-use script
5. Deploy it:
   • In Google Sheets, go to Extensions > Apps Script
   • Paste the code into a new script (name it, e.g., “Data Cleaning”)
   • Click “Run” and authorize permissions

Within seconds, your messy multi-tab file becomes a clean, consolidated dataset—fully automated.

Advanced Use Case from AI Finance Club

A member used this technique to:

1. Locate PDF invoices in Google Drive
2. Extract text/data from PDFs using AI
3. Auto-populate Google Sheets
4. Run financial analysis on the structured data

This would have taken days manually—now it’s a repeatable, one-click workflow.

4. Ask AI: “What Don’t I Know That I Don’t Know?”

This meta-strategy helps uncover blind spots in your business or role—especially valuable when starting a new job or scaling operations.

How to Set Up a Custom AI Assistant

The speaker uses a custom GPT loaded with:

• Personal bio
• LinkedIn profile
• Email templates
• Business data (e.g., SNOP—Sales and Operations Planning details)
• Admin/operational context

Then, ask: “What are 100 things I should know that I might not know?”

Surprising Insights AI Can Reveal

Examples from the speaker’s business:

• “What’s your exact total number of unique customers?”
• “Which product has the highest lifetime customer value?”
• “What’s the average time from someone following you to making their first purchase?”
• “How many hours monthly are spent on tasks automatable with Zapier?”
• “Do you have a clear beginner → intermediate → advanced learning roadmap?”
• “Could each course have a 30-second teaser video?”

Real Example: New Hire in Entertainment Park Industry

When an AI Finance Club member joined a theme park company, they asked AI:

“What are the unknown things I should know about running entertainment parks?”

AI revealed industry-specific blind spots:

• Cost per guest per attraction
• Regional pricing psychology (e.g., guests in Region X spend 22% more on snacks)
• Ride downtime = lost revenue (quantifiable!)
• Guest dwell time directly correlates with revenue
• Park cleanliness grades affect long-term attendance
• Cultural seasonality (e.g., local festivals drive spikes)
• Local GMs may prioritize volume over profit due to incentive structures

These insights would take months to discover organically—AI surfaced them instantly.

5. Leverage AI-Native Finance Tools: Puzzle & Concourse

For teams using modern financial infrastructure, AI-native tools automate accounting and reporting end-to-end.

Puzzle: AI-Powered Accounting for Startups

Ideal for: Startups using subledgers like Stripe, Gusto, Mercury, or Bill.

How it works:

• Connects directly to your subledgers
• Automatically posts transactions to your general ledger
• Uses AI to categorize, reconcile, and close books
• Generates real-time dashboards with KPIs (e.g., burn rate, CAC, LTV)

Result: Month-end close in minutes instead of days.

Concourse: AI Reporting for QuickBooks Users

Ideal for: SMBs using QuickBooks Online.

How it works:

• Connects to QuickBooks Online
• Automatically generates monthly management reports
• Compares actuals vs. prior month, budget, or forecast
• Allows you to edit text and visuals—AI regenerates the full report

Example output: A clean, narrative-driven report showing revenue trends, expense anomalies, and cash flow projections—all updated automatically.

Why AI-Native Tools Win

Unlike legacy software, tools like Puzzle and Concourse:

• Were built with AI at their core (not bolted on)
• Have no technical debt slowing innovation
• Rapidly integrate new AI capabilities (e.g., natural language queries, predictive alerts)

Note: The speaker confirms they are not affiliated with either company—they’re recommended purely based on product quality.

6. Use AI to Generate API Code for Automated Data Pulls

APIs let you pull live data from systems like QuickBooks, Salesforce, or HubSpot—but coding them used to require developer skills. Now, AI writes the code for you.

Step-by-Step: Automate QuickBooks Management Reports

Scenario: An accountant spends hours manually exporting P&L data from QuickBooks to Excel for reporting.

Prompt Strategy: Ask for Options First

Instead of jumping to a solution, use this prompt:

“I’m an accountant using QuickBooks Online. I manually build management reports in Excel, which takes hours. What are the different ways to automate this? List pros and cons of each.”

AI’s Recommended Solutions

After 55 seconds of research, ChatGPT provided six options:

1. Stay in QuickBooks: Use built-in reports (limited customization)
2. Add-on reporting tools: e.g., Fathom, Spotlight (costs $)
3. Excel add-ins: QuickBooks connector for Excel (basic)
4. Power Query + Power BI: Robust but steep learning curve
5. No-code automation: Zapier, Make, Power Automate (good for simple flows)
6. Custom API script: Full control, fully automated (best for complex needs)

Deep Dive: Generate the API Code

Choose option #6 and ask:

“Show me all steps to implement #6. I don’t know how to code—be super detailed.”

ChatGPT responds with a complete, beginner-friendly guide:

1. Prerequisites: QuickBooks Developer account, Python installed
2. Create a QuickBooks App: In developer.intuit.com
3. Prepare project folder: With config files for credentials
4. Install required libraries: pip install quickbooks
5. Run the script: A short Python code block that pulls P&L data

Sample Code Snippet (as generated by AI)

from quickbooks import QuickBooks
from quickbooks.objects import ProfitAndLossReport

client = QuickBooks(
    consumer_key='YOUR_KEY',
    consumer_secret='YOUR_SECRET',
    access_token='YOUR_TOKEN',
    access_token_secret='YOUR_TOKEN_SECRET',
    company_id='YOUR_COMPANY_ID'
)

report = ProfitAndLossReport.all(qb=client)
print(report)

Once set up, this script runs daily—no more manual exports. And since data is in Python, you can add forecasting, anomaly detection, or custom visualizations.

Choosing the Right AI Model Matters

Not all AI models are equal for finance tasks. The speaker emphasizes:

• Use GPT-4 (or “03”) for complex analysis—it takes longer (1+ minutes) but delivers deeper reasoning
• Enable “reasoning mode” or “thinking steps” to see how conclusions are reached
• For simple tasks (e.g., formula writing), GPT-3.5 or Copilot may suffice

Time invested in letting the AI “think” pays off in output quality—especially for KPIs, cohort logic, or script generation.

Security & Data Privacy Best Practices

When using AI with financial data:

• Never paste sensitive data into public/free AI tools
• Use enterprise-grade platforms (ChatGPT Teams, Copilot Pro) that guarantee data confidentiality
• If uploading isn’t allowed, use sample data (first 5 rows) to get logic/templates
• Run final code/scripts in your internal, secure environment

Real Results: Time Savings & ROI

The transcript highlights dramatic efficiency gains:

Tools Mentioned in This Guide

• ChatGPT (especially GPT-4/“03” model)
• Google Apps Script (for Sheets automation)
• Puzzle (AI accounting for startups)
• Concourse (AI reporting for QuickBooks)
• QuickBooks Online API
• Python (for data analysis and APIs)
• Zapier / Make / Power Automate (no-code workflows)

Who Should Use These Strategies?

These techniques are designed for:

• FP&A analysts
• Controllers & accountants
• Finance managers
• CFOs of SMBs and startups
• Finance team leads looking to upskill their teams

No data science or coding background required—just curiosity and willingness to experiment.

Getting Started: Your First AI Finance Project

Pick one strategy to implement this week:

1. Beginner: Use the KPI prompt on your latest P&L
2. Intermediate: Clean a messy dataset with AI-generated Google Script
3. Advanced: Set up an API pull from your accounting system

Start small, validate results, then scale.

Free Resource: AI for Finance Course

The speaker offers a free AI for Finance course (linked in the original video description) that teaches:

• All six strategies in depth
• Templates and prompts
• Security guidelines
• Real-world case studies

This is recommended for finance professionals serious about mastering AI quickly.

Final Thoughts: AI Is Your Finance Co-Pilot

The future of finance isn’t about replacing humans—it’s about augmenting judgment with speed and scale. As shown in this guide, AI can handle the repetitive, technical, and time-consuming tasks, freeing you to focus on:

• Strategic decision-making
• Stakeholder communication
• Business partnering
• Innovation

You don’t need a data scientist. You just need to start.

Best Ways Use AI in Finance: 6 Proven Strategies Every Finance Professional Can Implement Today

Buy this item