Understanding the New AI Assistants
The landscape of spreadsheet tools has evolved significantly. Instead of just writing formulas, you can now use AI assistants to perform tasks. These tools, such as Microsoft's Copilot for Excel and Anthropic's Claude, are designed to interpret plain-English
instructions. While some AI platforms focus on automating recurring reports by connecting to live data sources, spreadsheet-native assistants excel at helping you with the workbook you have open right now. They can generate formulas, build charts, and even structure entire models from a series of prompts, turning a conversation into a functional financial analysis.
Starting With a Clear Foundation
The quality of an AI-generated model depends entirely on the quality of your prompts and the data you provide. Before you begin, structure your assumptions clearly in an 'Inputs' tab. This is where you'll list key drivers like growth rates, costs, and inflation. Your first prompts should be simple and direct. For example, instead of a complex command, start by asking the AI to perform a discrete task: "Add a new column that calculates the year-over-year revenue growth." or "Summarize sales performance by quarter." This iterative approach allows you to build the model piece by piece and verify the AI's logic at each step.
Crafting Prompts for Dynamic Models
The real power of these tools is their ability to create dynamic models. This means building a spreadsheet where you can change a few key assumptions and have the entire model update automatically. To achieve this, you must instruct the AI to link its calculations back to your input cells. A powerful prompt might be: "Create a 3-statement financial model using the assumptions in the 'Inputs' sheet. The model should include scenarios for a base case, a bull case with 10% higher growth, and a bear case with 10% lower growth. Ensure all formulas are linked and dynamic." This instructs the AI to not only build the structure but also incorporate the sensitivity analysis that is crucial for strategic planning.
Best Practices for Effective Prompts
Writing effective prompts is a skill in itself. The key is to be specific, structured, and clear. Think of it as briefing a junior analyst. Start by defining the role you want the AI to play, such as, "Act as a senior financial analyst." Provide context, including the company, period, and any specific accounting standards. Finally, define the expected format, whether it's a table, a chart, or a full dashboard. For a complex request, break it down. You can even ask the AI to help you structure the main prompt itself before executing it. For example: "Create a detailed prompt I can use to ask Copilot to build a three-statement model with separate tabs for assumptions, income statement, and cash flow."
The Crucial Role of Human Oversight
AI is a powerful assistant, not a replacement for professional judgment. While these tools can build a model in minutes, they can also make mistakes or generate plausible-sounding but incorrect information. It is absolutely essential to review and audit the AI's work. Check the formulas, validate the logic, and trace key numbers back to their source. The primary role of the finance professional shifts from manual data entry to strategic review and interpretation. AI accelerates the 'what', freeing you up to focus on the 'why' and the 'what's next'. This combination of AI speed and human expertise leads to better, faster decision-making.
