Why Excel Breaks When FP&A Gets Serious

Your team built the model. Your team trusts the model. But has anyone checked the model?

If you deliver FP&A advisory services for a living, Excel is almost certainly the backbone of your practice. It is the tool you learned in, the tool your clients understand, and the tool that has powered financial modeling for decades. For simple analyses and one-off projects, it works fine.

The problems start when FP&A gets serious: when you are managing rolling forecasts across a dozen clients, when those forecasts need to satisfy bank covenants and board expectations, and when the margin for error narrows from “approximately right” to “defensibly accurate.”

At that point, Excel does not just become inconvenient. It becomes dangerous.

This is not a theoretical argument. The research is clear, and the real-world examples are sobering. Spreadsheet errors have caused billion-dollar losses, regulatory fines, and reputational damage at some of the most sophisticated financial institutions in the world. And if those organizations, with dedicated model risk teams and formal governance processes, cannot keep their spreadsheets error-free, the question for every accounting firm managing client models in Excel is: what makes you think yours are different?

The Scale of the Problem: What the Research Says

A 2024 peer-reviewed study found that 94% of business spreadsheets used in decision-making contain errors. The research covered a 35-year body of literature and was conducted in collaboration with Central Queensland University, Swinburne University of Technology, and City University of Hong Kong. This is not an outlier finding. It is a confirmation of what decades of study have consistently shown.

These are not typos or formatting issues. They are logic errors, broken references, hardcoded overrides, and formula inheritance problems that produce outputs everyone trusts, but no one has verified.

For an accounting firm managing ten client forecast models, the math is stark. If each model has a 94% chance of containing at least one error, the probability that all ten are clean is essentially zero. And unlike a single company’s internal model, your errors carry your firm’s name and your professional reputation.

Real-World Failures: When Spreadsheet Errors Make Headlines

The examples are not hypothetical. Some of the most expensive spreadsheet failures in history happened at organizations with resources, expertise, and controls that far exceed what any accounting firm has at its disposal.

These are not cautionary tales from decades ago. The Citigroup incident happened in 2020. The Public Health England failure happened during the COVID-19 pandemic. Spreadsheet risk is not a legacy problem — it is an active, ongoing source of financial and operational damage.

Why Excel Fails at Scale: The Five Breaking Points

Excel does not fail because it is a bad tool. It fails because it was never designed to be a multi-user, multi-model, continuously updated financial planning system. For accounting firms managing FP&A advisory across a growing client base, the breaking points are predictable and consistent.

1. Version Control Is Nonexistent

Every firm that has managed client models in Excel has experienced some version of this: two team members editing the same model, one saving over the other’s changes. Or a client referencing last month’s forecast file while your team has already moved to a revised version. Or a senior advisor making a “quick fix” directly in the model that no one else knows about.

Excel has no native version control. There is no audit trail of who changed what, when, or why. The undo history is limited and disappears when the file is closed. For a single user working on a single file, this is manageable. For a team of advisors maintaining a portfolio of client models, it is a constant source of risk.

2. Formula Errors Compound Silently

The most dangerous spreadsheet errors are the ones nobody notices. A formula that references the wrong row. A SUM range that does not extend to include a newly added line item. A hardcoded number that was meant to be temporary but was never replaced with a formula. A circular reference that Excel resolves through iteration instead of flagging.

In a simple model, these errors are relatively easy to catch. In a three-statement financial model with multiple tabs, driver-based assumptions, and linked schedules, they can persist for months or years. Each client model you build inherits the structural risks of every model before it, and every “just this once” shortcut compounds over time.

3. Multi-Client Scaling Is Manual and Fragile

When an accounting firm starts FP&A advisory with two or three clients, Excel works. Each model is bespoke, each advisor knows every cell, and the personal attention keeps quality high.

At ten clients, the cracks appear. At twenty, the system is held together by institutional knowledge and heroic effort. Every client model is slightly different — different tab structures, different naming conventions, different formula approaches. When the advisor who built a model leaves the firm, knowledge walks out the door. When a new team member takes over, they spend weeks understanding the model before they can safely update it.

There is no way to push a structural improvement across all client models simultaneously. If you discover a better way to model revenue recognition or working capital, you must manually rebuild that logic in every workbook. This is not a technology limitation — it is a fundamental architectural constraint of file-based modeling.

4. Assumptions Are Buried, Not Explicit

In a well-built Excel model, assumptions are separated on a dedicated tab and clearly labeled. In practice, most models are not well-built. Assumptions get embedded in formulas, hardcoded in cells, and scattered across tabs. The result is a model where the outputs look reasonable but the reasoning behind them is opaque.

This matters enormously when forecasts are used for consequential decisions. A bank reviewing a loan application needs to understand the assumptions behind a cash flow projection. A board evaluating an expansion plan needs to know which revenue growth rate the model assumes and why. When assumptions are buried in cell formulas rather than surfaced as explicit, adjustable inputs, the model becomes a black box that produces numbers without explanation.

5. Compliance and Audit Risk Is Growing

Regulatory scrutiny of spreadsheet-based financial processes has increased significantly. The Basel Committee on Banking Supervision has issued guidance specifically addressing spreadsheet risk in financial institutions. SOX compliance requirements increasingly extend to the tools used to prepare financial reports.

For accounting firms, this translates into professional liability. If a client’s bank covenant calculation is wrong because of a spreadsheet error, or if an investor presentation contains a forecast built on a broken formula, the firm that built the model bears responsibility. The lack of an audit trail, version history, or structural validation in Excel makes it difficult to demonstrate the rigor that professional standards demand.

The Pattern: When Excel Stops Being Enough

Most accounting firms do not hit all five breaking points at once. The progression is predictable:

• Phase 1: The model works. You built it, you understand it, you trust it. One to three clients, one advisor per model.
• Phase 2: You start copying models for new clients and modifying them. Small inconsistencies creep in. An occasional version control issue, but nothing that causes a visible problem.
• Phase 3: Team members inherit models they did not build. Understanding what a model does takes longer than updating it. You start finding errors that have been present for months.
• Phase 4: A client or their lender asks a question that reveals an error or an assumption you cannot trace. The model is correct “enough” but not defensibly accurate. Your confidence in your own work starts to erode.
• Phase 5: You realize that the time spent maintaining, fixing, and explaining Excel models is the same time you should be spending on the advisory work that actually creates value for clients.

The firms that move beyond Excel are not the ones that hate spreadsheets. They are the ones that have hit Phase 4 or 5 and recognized that the risk and inefficiency of spreadsheet-based modeling is actively limiting their ability to scale.

What “Moving Beyond Excel” Actually Means

Moving beyond Excel does not mean abandoning analytical rigor. It does not mean learning an entirely new way of thinking about financial models. And it does not mean losing the flexibility that made spreadsheets useful in the first place.

What it means is replacing the fragile, uncontrolled, file-based workflow with a purpose-built platform that provides the same modeling capabilities — driver-based assumptions, three-statement output, scenario analysis — in an environment with version control, audit trails, multi-user collaboration, and the ability to standardize and scale across a client portfolio.

The financial model itself does not change. The methodology does not change. What changes is the infrastructure underneath it — from a collection of independent files that depend on individual knowledge and discipline to a structured system that enforces consistency, catches errors, and makes the work repeatable.

For accounting firms managing FP&A advisory across multiple clients, this is not an upgrade. It is a risk management decision. The same way your clients moved from desktop QuickBooks to cloud accounting, the move from Excel to a purpose-built FP&A platform is about eliminating a category of risk that has been tolerated for too long because there was no practical alternative.

Ready to eliminate spreadsheet risk from your advisory practice? See how Jirav replaces Excel financial models with a platform built for multi-client FP&A. Request a Demo here