In Polish companies — from startups to large agricultural enterprises — three fundamentally different management tools are often conflated: the budget, the financial model, and the activity-based financial-operational model. This is not a matter of semantics: it is the difference between a tool that says “how much we spent” and one that answers “why we spent that much — and what happens if I change one operational decision tomorrow.”
This article explains the differences using a concrete example: FarmWise — an MFO for a dairy farm created by Origami Effect in 2024. It is the company’s 190th financial-modelling implementation: over 3 million cells and ~300 sheets simulating every aspect of a dairy cow’s life and production.
Budget — an expense-control tool
A budget is a financial plan that sets how much will be allocated to specific spending categories over a period — usually a year. It is a control instrument: it allows comparing actual costs against plan and detecting variances.
Example questions a budget answers:
- How much did we plan to spend on feed in Q1?
- Did we exceed the veterinary budget?
- What are revenues vs plan?
The fundamental limitation of a budget
Budgets are typically static — built once at the start of the year using historical data and judgment. They do not simulate cause-and-effect relationships: the “feed” category remains a single aggregated number, with no breakdown explaining its components.
Budgets draw data from accounting inventories (herd size at period start/end), which suffices for simple valuation and control but not for production simulation (e.g., how changing a feed recipe affects cost per liter of milk).
Financial model — a dynamic projection
A financial model is a system of linked formulas that represent the company’s financial structure over time. Unlike a budget, it is dynamic: changing one assumption propagates through the P&L, balance sheet and cash flow.
Typical questions a financial model answers:
- How does a change in interest rate affect profitability?
- What is the DCF valuation under a pessimistic scenario?
- When will we reach break-even?
How formulas work in a financial model
Formulas in a financial model link financial variables and propagate changes. Example: monthly milk revenue = milk production × average farmgate price. A price change of a few cents automatically recalculates revenues, margins and cash flow for the whole year.
However, financial models usually aggregate data at the financial-category level and do not directly describe operational processes.
Activity-based model (MFO) — simulating reality
An MFO (Financial-Operational Model) starts from activities and operational processes — not aggregated financial categories. Every cost and revenue is the result of formulas that describe what physically happens in the business (e.g., how many cows are in a given lactation phase, how many kg of a specific feed ingredient are consumed, etc.).
In an MFO the formulas describe reality, and the financial outcome is its consequence — the inverse logic of a budget.
Example questions an MFO answers
- What does it cost to produce one liter of milk after changing the feed recipe?
- How does an increase in mastitis incidence by 3 percentage points affect production and margin?
- Is it better to vaccinate prophylactically or treat reactively (including production losses)?
- How much straw must we buy in Q3 given the herd structure?
FarmWise — how formulas simulate herd life
FarmWise is a practical MFO implementation: over 3 million cells and ~300 sheets. Below are key mechanisms illustrating the difference between MFO, a budget and a classic financial model.
Herd-structure simulation
FarmWise does not treat the herd as a single number. Each animal group has assigned age and cycle stage (calf, heifer, milking cow, dry) and lactation stage (1st–7th cycle). The model tracks calving dates, inseminations and probabilistic losses, so it knows the future herd composition.
The Dairy Cycle module defines average daily milk yield for each lactation cycle; Milk Production multiplies these yields by the number of cows in each group (taken from Inventory) and calculates monthly/quarterly production.
Detecting health deviations in the herd
The model computes disease risk per condition and age group (mastitis, ketosis, lameness, BVD, leptospirosis). Multiplying probability by group size produces a forecasted case count — enabling early detection of deviations and preventive responses.
The model compares strategies: reactive treatment vs preventive program (e.g., therapy cost × case count vs vaccination cost × number of animals), enabling data-driven choices.
Precision in feed costs and recipe changes
Users can define up to 10 feed recipes (each with an effective period); each recipe contains kg per cow per day for individual ingredients (silage maize, alfalfa, molasses, etc.).
The model multiplies daily intake by herd size and days in the quarter — where herd size is the result of biological simulation, not a manual input. Thus feed demand and costs always stem from the herd state and forecast.
The model distinguishes between self-produced and purchased feeds, accounts for storage losses, and calculates the cost of producing on-farm feed from scratch (fuel, fertiliser, machinery depreciation, land financing, etc.).
Cascading changes when a recipe is modified
A recipe change (e.g., 2 kg alfalfa → 3 kg silage per cow in a 300-cow herd) immediately affects silage demand, the need to purchase raw material, cost per liter of milk, EBIT and cash flow — all automatically and without manual recalculation.
Comparison: budget, financial model and MFO
| Dimension | Budget | Financial model | Activity-based model (MFO) |
|---|---|---|---|
| Main purpose | Expense control | Projection & valuation | Operational simulation → financial outcomes |
| Starting point | Cost categories | Financial assumptions | Physical activities (cow, lactation, recipe) |
| Formulas | Simple sums | Complex financial links (DCF, WACC) | Multi-layered: biological and financial together |
| Granularity | Category / department | P&L lines, balance, CF | Each cow, each feed ingredient, each field |
| Scenarios | None | Optimistic/base/pessimistic | Unlimited: recipe, disease, machine |
| Variance detection | Ex post | At assumption level | Proactively — before the result shifts |
| Cost per liter of milk | Aggregated, historical | Forecasted by category | Calculated from each ingredient |
| Feed-recipe change | No effect | Changes cost category | Cascade: recipe → cost → EBIT |
Formulas as the foundation of every model
All three tools rely on formulas: data are inputs, and formulas define how to turn those inputs into decision-ready knowledge. The difference is what the formulas describe:
- Budget: category sums.
- Financial model: financial linkages and propagation through P&L/balance/CF.
- MFO: biological/agronomic processes; the financial result follows the operational simulation.
In FarmWise the cost per liter of milk is the outcome of hundreds of formulas — from the field, through the barn, to the income statement.
When to use which instrument?
Each tool has its purpose:
- Budget — annual control of cost-plan execution.
- Financial model — raising finance, DCF valuation, enterprise-level scenario planning.
- MFO — decisions dependent on operational processes: recipe optimisation, veterinary strategy, crop planning for feed needs, machinery profitability analysis.
The budget asks about limits. The financial model asks about projections. The MFO computes possibilities.