Claude using Excel: 5 Real-world Financial Modelling Use Cases

I’ve found that Claude using Excel isn’t just about building faster models.

The real win is how it helps me structure the work: clarifying questions first, assumptions second, and then connecting the logic across statements so the model stays coherent instead of becoming a debugging exercise.

Below are Five real-world FP&A financial modelling use cases I used as templates for building complete Excel models with Claude. I’ll also share the modelling mindset behind each one, so the output is useful even if you rebuild it manually later.

What “good architecture” means in Financial Modelling

A three-statement model is not three separate reports. It’s a connected system.

Net income flows into retained earnings.
Depreciation impacts both income statement and cash flow.
Working capital changes quietly control operating cash flow.
Debt affects interest expense and financing cash flows.

Subscribe to our YouTube Channel for Free Videos on FP&A. Click here to Subscribe.

Follow along with step-by-step examples, videos, and ready-to-use datasets here: https://fpnaprofessionals.teachable.com/l/pdp/claude-with-excel 🚀

Use case 1: Build a Complete Three-Statement Financial Model

My starting point for Claude using Excel was a full restatement for a “new college setup” style business. The goal was to generate:

Income statement (P&L)
Balance sheet
Cash flow statement
All linked dynamically through time

How Claude using Excel accelerates the process

Instead of me writing every assumption upfront, Claude asks clarifying questions first. For the college example, it questioned items like:

How student growth should behave beyond initial intakes (growth continuation or cap)
Tuition fee per year (a pricing assumption I initially hadn’t specified)
Projection horizon (5 years, 7 years, or 10 years)
Funding capital structure (equity-only vs debt/equity mix)
Campus approach (lease vs purchase) and related currency assumptions

Then it executes in phases:

Create the assumptions sheet
Build the income statement
Build the balance sheet
Build cash flow and link ending cash back to the balance sheet
Run tie-out checks

The important part is not speed alone. It’s that Claude forces the model to be consistent across statements. That reduces the most painful FP&A failure mode: sign errors and circular logic that break cash reconciliation.

Use Case 2: Evaluate a SaaS pricing restructuring with real assumptions

Next I used Claude using Excel for a revenue forecasting decision: switching from a single-tier subscription to a three-tier pricing structure.

Economics I started with

1,000,000 new registrations per month
10% month-over-month growth in registrations
85% active users
2% paid conversion from active users
$19/month paid subscription in the single-tier model

What I asked Claude to build

I prompted Claude to generate a 24-month monthly revenue forecast under two scenarios:

Single tier subscription
Three-tier structure with different plan prices (basic, professional, professional plus)

Then I required outputs that support CEO-level decisioning:

New users, active users, paid users per month
MRR and total revenue per month
A comparison tab showing revenue and MRR differences
An executive summary with a data-driven recommendation

This is a great pattern for pricing work: treat it like an economics model, not a spreadsheet template. Claude’s value comes from structuring the scenario logic and keeping the metrics consistent.

Use Case 3: Scenario Panning (Base, Best, Worse) with a Dynamic Switch

Scenario planning answers a different question than sensitivity analysis. Sensitivity changes one variable. Scenario planning changes multiple assumptions together to represent plausible futures.

My approach is typically:

Base case: current assumptions
Best case: higher revenue with higher volumes
Worst case: lower revenue with lower volumes

Traditional Excel approach (the “choose + dropdown” pattern)

In pure Excel, I’d usually combine:

CHOOSE to select which scenario values to use based on an index number
A combo box dropdown to set that index number interactively

How Claude using Excel does it faster

With Claude using Excel, I asked it to perform scenario analysis and build:

A dynamic scenario switch (dropdown)
formula wiring to update results
optional dashboard-like summary structure

The key win here is that Claude reads the dataset structure, understands the relationship (like revenue = price x volume), and then wires the scenario output to a selectable switch.

Use Case 4: Sensitivity Analysis Tables (Excel techniques vs Claude automation)

Sensitivity analysis asks: What happens if this changes?

In practice, it usually means altering one or two assumptions like:

Price
Volume
Cost

And observing the impact on outcomes like revenue or profit. If small changes create big movement, that variable is “high sensitivity” and represents key risk or opportunity.

Example Model: Website Ad revenue

I used a website revenue setup based on these drivers:

Monthly visitors
Clickthrough rate
CPM and CPC style metrics

The baseline output was computed, then a sensitivity table tested changes in:

Visitors (for example, 750k to 1M)
Clickthrough rate (for example, 2% to 2.5%)

Traditional Excel method

Typically this becomes a What-If Analysis data table setup, where you define:

Row input cell
Column input cell

It works, but it’s another place where you lose time learning the mechanics and debugging broken references.

Claude using Excel method

I prompted Claude to update the sensitivity analysis table using the dataset. It:

read the inputs
identified the sensitivity table structure
updated the formulas accordingly
reflected the new outputs in the grid

The outcome is the same, but with less friction and more time spent interpreting what the sensitivity actually means.

Use Case 5: Working Capital Forecasting

Working capital looks simple on paper: current assets minus current liabilities.

But FP&A reality is messier. Working capital is where cash gets delayed, blocked, or preserved.

My definition that keeps cash flow honest

Receivables increase means cash is delayed
Inventory increases means cash is blocked
Payables increase means cash is preserved

Traditional driver-based approach

I’d normally forecast by computing and projecting ratios like:

Days Sales Outstanding (DSO) for receivables
Days Inventory Outstanding (DIO) for inventory
Days Payable Outstanding (DPO) for payables

Then I reverse-calculate account balances for each year, compute changes, and link to cash flow. The catch is sign correctness. Get the sign wrong and cash flow can flip direction.

Claude using Excel automation

With Claude using Excel, I prompted it to build a 5-year working capital forecast from historical account balances. Claude:

Read existing historical data
Checked dataset labels/columns
Calculated historical “days” ratios
Projected those days forward
Reverse-calculated receivables/inventory/payables balances

The payoff is less manual wiring and fewer sign/ref mapping errors, while keeping the driver logic clear.

My Practical Takeaway: Use Claude to remove friction, then focus on judgment

My main message is simple: AI doesn’t replace strategic thinking. It compresses the tedious parts so what remains is the part that matters most in FP&A: your assumptions, your interpretation, and your ability to explain what the numbers mean.

If you want to explore more, I’d encourage you to start with one model type that you already build every month. Use Claude using Excel to wire it correctly end to end, then iterate on the assumptions until the model tells a story you trust.

I encourage you to experiment with these prompts and keep the focus on interpretation rather than mechanics.

FAQ’s

Q1. Is Claude using Excel only useful for speed, or does it improve model quality?

It improves both. The key quality lift comes from how Claude structures assumptions, asks clarifying questions, and connects logic across statements. That reduces tie-out failures like cash not reconciling between cash flow and the balance sheet.

Q2. What’s the biggest Modelling Risk when building Financial Models Manually?

Operational friction that turns into errors: circular references, sign mistakes in working capital to cash flow, and broken architecture where statements don’t reconcile.

Q3. How do Sensitivity Analysis and Scenario Planning differ?

Sensitivity analysis changes one or two variables to see impact on results. Scenario planning changes multiple assumptions together to represent different plausible futures like base, best, and worst cases.

Q4. What inputs should I provide Claude using Excel for a pricing forecast?

You should provide driver economics (registration growth, activation rate, conversion rate, and pricing per tier), plus the time horizon. If you want decision-ready output, request MRR, paid users, revenue by month, and an executive comparison summary.

Q5. Does AI replace my finance skills?

No. AI reduces the operational work (debugging formulas, wiring scenario switches, building tables). Your judgment still determines which assumptions are realistic and how you interpret the model outputs.

Follow along with step-by-step examples, videos, and ready-to-use datasets here: https://fpnaprofessionals.teachable.com/l/pdp/claude-with-excel 🚀

Join the FP&A Community and Keep Learning

If you are serious about FP&A then checkout CGFPA® Certified Global FP&A Professional Program. A six-month in-depth CPD accredited FP&A Certification Download the CGFPA® Brochure.

You can find more information and bunch of free FP&A resources, including books, courses, and free templates, on the FP&A Professionals Institute website www.fpnaprofessionals.com

Subscribe to Weekly FP&A Jobs Newsletter (every Wednesday)

Subscribe to our YouTube Channel for Free Videos on FP&A. Click here to Subscribe.

Find out about our Weekly Live Webinars: https://fpnaprofessionals.com/