The annual budget was a sensible invention for its time: a factory that knows what it will produce in January can reasonably plan its costs through December. In 2026, that model freezes an organisation for twelve months while energy prices, supplier lead times, and interest rates can reverse course in a matter of weeks.
Two methods have emerged to replace — or complement — the annual budget: rolling forecast (rolling projections over 12 to 18 months) and zero-based budgeting (every budget line starts from zero each cycle). Neither is an accounting revolution. They are different ways of exploiting the data your ERP is already collecting.
This guide explains how to implement them in practice, which tool to choose based on your company’s size, and which pitfalls to avoid so the project does not simply become a new Excel nightmare.
Rolling Forecast: Principles, Cycle, and ERP Implementation
What Is a Rolling Forecast?
A rolling forecast is a budget that moves forward in time. Instead of locking an end-of-year horizon and holding to it, you permanently maintain a projection 12 or 18 months out from today.
A concrete example: in July 2026, your forecast horizon runs from July 2026 to June 2027. At the end of July, you update the model and the horizon slides to August 2026 – July 2027. Each month, you add one month of projection at the far end, and you refine the next three months using actuals that have just entered the ERP.
The fundamental difference from the annual budget:
- Annual budget: frozen from the previous November, operational through December of the target year. Any deviation from actuals is a variance to explain, not a signal to incorporate.
- Rolling forecast: updated each month. The question is not “why did we miss the budget?” but “what does our 12-month projection look like given today’s data?”
This shift in mindset is as cultural as it is technical: the rolling forecast assumes that actuals carry more value than the original target.
Configuring Rolling Forecast in Your ERP
The reality of mid-market ERPs is nuanced: most can store budgets and compare actuals to forecasts. Few genuinely handle a native rolling forecast — versioning rolling projections, managing multiple scenarios, and consolidating decentralised contributions.
SAP S/4HANA: the CO (Controlling) module manages cost centres and budgets. For true rolling forecast functionality, SAP recommends SAP Analytics Cloud (SAC), which connects to the ERP core in real time. The forecast layer lives in SAC; the ERP supplies the actuals.
Odoo: Odoo’s native Budget module lets you enter budgets by analytical account and track variances in real time. It does not natively support rolling revisions. Most mid-market companies using Odoo connect an external FP&A tool (Causal.app, Pigment) via Odoo’s REST API for rolling forecast.
Sage X3 / Sage 200: both products offer a solid budgeting module for actual-vs-budget comparison. Rolling forecast is typically handled through the integrated BI layer (Crystal Reports or Power BI) fed by scheduled extracts. Functional, but largely manual.
Microsoft Dynamics 365 Finance: native Budget Planning with a Power BI connector included in the licence. Microsoft recommends Power BI or Azure Synapse Analytics for complex rolling projections. Connectors are well documented.
Summary: beyond £30–40M revenue, most finance teams quickly hit the ceiling of their ERP’s native budget module and start looking for an EPM or FP&A overlay. Below that threshold, a well-structured monthly export to Sheets or Excel can be sufficient.
Key Drivers to Manage in Your Rolling Forecast
A useful rolling forecast does not model every line of the income statement. It is built around drivers — key variables that explain the majority of your costs and revenues.
Drivers to connect directly to your ERP:
- Sales pipeline (integrated CRM): conversion rate × average deal value → revenue forecast
- Order book: confirmed orders not yet shipped → near-certain short-term revenues
- Inventory levels and supplier lead times: early warning signal for supply constraints
- Forward payroll: planned headcount additions + salary reviews + extended leave → staff costs for N+1 to N+6
- Forecast EBITDA by profit centre: aggregation of drivers by analytical dimension
A driver-based rolling forecast is structurally less precise than a line-by-line budget over the next 12 months — and deliberately so. It is designed to capture trends, not to reproduce the accounting of the future.
Zero-Based Budgeting: Principles and Use Cases
What Is ZBB?
Zero-based budgeting starts from a simple premise: a spend item is not justified simply because it existed the previous year. Every budget line starts from zero. Every department head must explicitly request and justify every pound spent, instead of benefiting from an automatic rollover at +2% or +3%.
Popularised in the 1970s — Jimmy Carter applied it to Georgia’s state administration before becoming US President — then adopted widely by consulting firms and private equity funds in the 2010s and 2020s, ZBB is above all a transformation tool, not a recurring management tool.
That distinction matters: the rolled-over annual budget is comfortable but accumulates obsolete spend. The rolling forecast optimises the trajectory. ZBB resets the cost structure itself.
ZBB in Your ERP: Configuration and Limits
From a technical standpoint, ZBB in an ERP works as follows:
- Clearing prior-year budgets: at the start of the cycle, prior-year budget lines are archived but not copied forward to the new financial year. Managers begin with a zero budget in every cost centre.
- Entry and justification: each manager submits budget requests with a rationale for each line. Most ERPs allow a comment or supporting document to be attached to a budget line.
- Prioritisation and arbitration: requests are aggregated and presented to the executive committee for decision. This is the most time-intensive phase — typically two to three times longer than a standard rollover.
- Validation and upload: approved budgets are loaded into the ERP for monthly actual tracking.
Tools that streamline this workflow: Anaplan, Workday Adaptive Planning, and Oracle EPM Planning each offer dedicated ZBB modules with built-in approval workflows. The ERP alone can manage steps 1 and 4, but steps 2 and 3 (request collection, prioritisation) generally require a collaborative external tool or custom-built forms.
Primary limitation: ZBB is resource-intensive. In a mid-market business of 200 to 500 employees, a full ZBB cycle absorbs three to six weeks of finance and operational effort. That is acceptable for a post-acquisition transformation; it becomes a penalty as a systematic annual process.
When Should You Apply ZBB?
ZBB does not fit every situation. It is appropriate in three specific contexts:
Post-LBO (private equity transaction): PE funds frequently impose ZBB in the first 12 to 18 months following an acquisition to identify structural costs that the previous management team never questioned.
Post-merger integration: when two companies merge, cost duplications are difficult to detect under a rollover budget. ZBB forces analysis of each line item in the light of the new consolidated entity.
During a crisis or transformation plan: when a business must fundamentally reassess its priorities — strategic pivot, intense competitive pressure, restructuring — ZBB provides a framework to legitimise difficult trade-offs.
Outside these situations, ZBB is disproportionate. The majority of SMEs and mid-market companies gain more from combining a rolled-over annual budget with a monthly rolling forecast than from running ZBB every year.
ERP Alone or ERP + EPM: Which Setup for Your Size?
This question comes up consistently in FP&A modernisation projects: should you buy a separate EPM tool, or extract maximum value from what the ERP already offers?
| Company profile | Recommendation | Rationale |
|---|---|---|
| SME, revenue < £20M, finance team of 1–3 | ERP native + structured export (Sheets or Excel) | An EPM tool (£10k–£40k/year) costs more than the value it adds for a simple structure |
| Mid-market, £20M–£150M, dedicated financial controller | ERP + lightweight FP&A tool (Causal, Pigment, Abacum) | Modern FP&A tools offer 12–18M rolling forecast, scenario planning, and visualisation from ~£400–£1,500/month |
| Upper mid-market, revenue > £150M, multi-entity consolidation | ERP + structured EPM (Workday Adaptive, Oracle EPM, Anaplan) | Multi-currency consolidation, intercompany eliminations, and IFRS requirements justify a dedicated EPM |
| Enterprise, > £400M, complex structures | SAP Analytics Cloud / Oracle EPM Fusion integrated with ERP | Simultaneous RF and ZBB, balance sheet stress testing, and long-range planning require an enterprise EPM platform |
The standalone FP&A tool market evolved significantly in 2024–2025. Players such as Pigment, Causal, Abacum, and Mosaic have captured a meaningful share of mid-market companies seeking something more accessible than traditional EPM suites (SAP, Oracle, IBM) without committing to spreadsheets. These tools connect to virtually all major ERPs via API or native connectors.
Practical Case: Implementing a Rolling Forecast in Six Weeks
Here is a realistic implementation plan for an SME or mid-market company starting from an annual Excel budget and looking to move to a monthly rolling forecast within its ERP — or a tool connected to it.
Week 1: Map your cost centre structure
Identify the analytical dimensions already set up in your ERP. If the analytical plan is outdated or too granular, now is the time to simplify it. Aim for 10 to 25 cost centres to start. More is not better: every dimension added multiplies entry and analysis complexity.
Week 2: Define the horizon and frequency
Choose 12 or 18 months. For a first implementation, 12 months is more realistic: forecast accuracy falls away sharply beyond that, and operational teams struggle to project 18 months out. Decide on the update frequency: monthly is the standard; quarterly is acceptable for less mature organisations.
Week 3: Extract and clean historical data
Extract three years of actuals from the ERP (costs, revenues, key volumes) by cost centre. This step often reveals inconsistencies in the analytical plan: misposted charges, cost centres that have been empty for months, reclassifications that were never reflected. Clean before modelling.
Week 4: Build the driver-based model
Do not reproduce the income statement line by line in your rolling forecast model. Identify the 5 to 10 variables that explain 80% of your costs and revenues: order volume, average headcount, key input prices, CRM pipeline conversion rate.
Week 5: Connect operational data sources
Plug in the CRM (sales pipeline), payroll (forward staff costs), and ERP purchasing data into your model. If you are using an external FP&A tool, configure the connectors at this stage. If you are staying on the native ERP, set up automated extracts.
Week 6: First executive committee validation cycle
Present the first rolling forecast to the executive committee not as a fixed budget but as a navigation signal. Show the gap between the initial budget and the updated RF. Adjust the model based on operational feedback. The first cycle is for calibration, not perfection.
5 Common Mistakes to Avoid
1. Projecting 24 months from day one
A 24-month horizon creates the illusion of precision while demoralising the operational teams who must populate figures for such a distant future. Start at 12 months. You can always extend the horizon once the model is bedded in.
2. Copying the annual budget structure into the rolling forecast
The temptation is strong: recreate the same table as the annual budget but with rolling monthly columns. This is the worst way to do rolling forecast — you keep all the complexity of the annual budget while changing none of its logic. RF demands a driver-based approach, not a future-dated accounting exercise.
3. Neglecting data collection automation
If your finance team spends two days every month manually extracting data from the ERP, pasting it into Excel, and reformatting it, the rolling forecast will quickly become a burden that nobody maintains seriously. Data feed automation is a non-negotiable condition for a sustainable process.
4. Excluding operational managers from the process
A rolling forecast built solely by the finance function from ERP figures misses the point. Sales managers know which deals are in active negotiation; production managers see supply risk building. Without them, the RF is retrospective, not prospective.
5. Confusing rolling forecast with sales forecasting
The sales forecast is an input to the rolling forecast, not the rolling forecast itself. A complete RF also incorporates variable costs linked to sales, fixed overheads, planned capital expenditure, and cash flows. Several SMEs have implemented what they called a “rolling forecast” that was, in practice, little more than an enhanced sales pipeline.
Rolling Forecast and ZBB: Two Complementary Tools
Rolling forecast and zero-based budgeting are not in competition — they operate at different points in the financial cycle.
ZBB is a reset tool: applied selectively to rebuild a company’s cost structure, typically following a major transformation or capital event. It produces a robust reference budget, defended line by line.
The rolling forecast is a continuous navigation tool: once the reference budget (ZBB-derived or otherwise) is set, it enables monthly forecast adjustments as conditions change.
The ideal sequence: apply ZBB to construct the budget for the following financial year during a transformation, then manage execution of that budget via a monthly rolling forecast. The ERP alone rarely handles both simultaneously beyond £150M revenue. An EPM or FP&A overlay quickly becomes essential.
Download our ERP evaluation scorecard: 30 criteria across 100 points to benchmark three vendors side by side, including their native FP&A and budgeting coverage.
To go further, read our complete guide to ERP vs EPM and FP&A integration and our article on management accounting and budget control in your ERP.
