7 Financial Modeling Best Practices Every FP&A Professional Must Know

If you have spent any time working in Finance and FP&A, you have at least once opened a model that made no sense.

Not because the data was wrong, but because whoever built it knew something you did not and left no trail. Financial modeling sits at the centre of everything FP&A does, from budgets and forecasts to long-range planning, and yet most Finance teams are still building models that only one person truly understands.

For years, FP&A teams have patched these problems with Excel workarounds: colour-coding conventions, tab naming rules, manual check formulas, and version-numbered files saved across shared drives. These patches work, up to a point. But they rely entirely on human discipline. The moment someone deviates from the convention, the model breaks.

Farseer was built to solve this at the platform level. Rather than asking finance teams to follow conventions, it makes the right structure the only option.

This article walks through seven financial modeling best practices and shows how a modern FP&A platform handles each one, replacing the manual Excel workarounds most teams still rely on today.

In Excel, formatting is a pain. Someone on the team decides on a colour-coding convention blue for inputs, black for formulas, green for cross-sheet links. And then everyone has to remember it, follow it, and enforce it manually. The moment a new analyst joins, or someone works late and stops caring, the convention breaks.

The deeper problem is that Excel relies on conventions to distinguish between an input cell and a formula cell at the structural level. They look identical unless someone has formatted them differently. A hardcoded number in a formula column is invisible until the model breaks.

Nothing is worse than hardcoded numbers buried inside Excel formulas. 

A formula like =B5*1.05 isn’t a model. It’s a guess. A model is dynamic. It allows you to change the growth rate in one place and watch the impact ripple through revenue, costs, and cash flow instantly. 

That dynamism only works when every assumption lives in a dedicated, clearly labelled assumptions sheet.

The assumptions sheet is the control room of your model. It is the single place where a CFO, an auditor, or a new team member can go to understand what the model believes about the future. 

Every macro input — growth rates, margins, FX rates, inflation, headcount targets — should sit here with its source, owner, and last-updated date clearly documented.

Read: What Is Margin Analysis and How to Conduct It Step by Step

Picture one enormous Excel sheet with revenue, costs, P&L, balance sheet, and scenarios all tangled together. That is the huge Excel monster most finance teams end up with. One broken link or a missed reference cascades everywhere. Troubleshooting takes hours because there is no clear separation between what feeds what.

In 2012, a modeling error in JPMorgan Chase’s Value-at-Risk model contributed to what became the London Whale trading loss and eventually over $6 billion. 

The model had a formula that used the average of two prices instead of their sum. One formula. Billions of dollars. Financial models carry real stakes. Error checking is not optional.

In Excel, error checking is something you build on top of your model: a dedicated checks tab, IF (condition, “PASS”, “FAIL”) formulas, conditional formatting to turn cells red when something breaks. It works, but it is entirely manual. You have to build it, maintain it, and remember to look at it.

A single-point forecast is not a financial model. The future is uncertain, and FP&A’s job is not to predict what will happen but to help leadership understand the range of outcomes and make better decisions across that range. Scenario and sensitivity analysis are the tools that deliver this.

In Excel, running three scenarios means maintaining three versions of the same model or building a scenario selector with CHOOSE () logic and a dropdown. Either way, the moment you add a fourth scenario or change a shared assumption, you are manually updating multiple places. The risk of inconsistency is constant.

Undocumented models are black boxes. They cannot be audited. They cannot be handed over. They cannot be updated by anyone other than their original builder. And even that person sometimes forgets the logic after three months away from the file. Documentation is not administrative overhead; it is part of the model itself.

In Excel, documentation is something you build separately: a cover sheet, inline cell comments, a changelog tab, source columns next to assumption cells. All of it requires discipline and maintenance. None of it is enforced.

A budget model built in October is not the same model in March. Actuals have come in. Assumptions have been revised. The board approved a new headcount plan. New product lines were added. A model that is not systematically maintained drifts from reality — and once it drifts, it becomes worse than no model at all.

In Excel, version control is a file naming convention: ModelName_v1.0_YYYYMMDD_Initials.xlsx. It works until it does not — until someone saves over the master, or five versions of the file circulate simultaneously with no clear source of truth.

Financial modeling is a craft. The teams that invest in building models the right way, rather than just the fast way, are the ones whose analysis gets trusted, acted on, and remembered. Farseer does not change the craft. It removes the friction that has always stood between knowing the best practices and actually following them.