.svg)
Forecasting for CFOs: budget, rolling forecast, latest estimate and 13-week cash flow (2026)
Four forecast types, one principle: your model belongs in a spreadsheet. Here's what each forecast is for, when to run it, and how to stop treating actuals as a manual chore.
.png)
SME CFOs run four distinct forecasts, not one: budget (annual), rolling forecast (monthly, pragmatic), latest estimate (half-yearly, deep — usually for PE investors), and 13-week cash flow (weekly, cash-based). The first three are accrual variants at group, entity, or cost-center level; the 13-week cash flow runs separately. The model belongs in a spreadsheet because real forecast architecture is too varied for any fixed UI. Finstack pushes consolidated actuals straight into that spreadsheet, closing the loop without locking you out of it.
Forecasting for CFOs: budget, rolling forecast, latest estimate and 13-week cash flow (2026)
A practical guide to the four forecast types, the dimensions that qualify them, and why your model belongs in a spreadsheet.
TL;DR:
Modern SME CFOs run four forecasts, not one. The budget is your annual commitment to investors and the board. The rolling forecast is your operational steering instrument — 12-18 months ahead, refreshed every month, fast and pragmatic, common with venture capital investors. The latest estimate is a deeper half-yearly rebuild of the budget for the remaining months, common with private equity. And the 13-week cash flow forecast is your weekly liquidity tool — cash-based, not accrual. Budget, rolling, and LE share the same accrual mechanics and can run at group, entity, or cost-center level. The 13-week runs separately. Keep the model in a spreadsheet: tool UIs can't handle the variation in how real businesses model revenue and cost, and you'll end up back in Excel anyway. Use a tool for the data layer: consolidated actuals on tap, automatic refresh into your sheets, forecasts written back to the system in one click. That's the combination Finstack runs — consolidation and forecasting in the same environment, without forcing you out of the spreadsheet where the modelling actually happens.
What is forecasting for SME CFOs?
Forecasting projects your P&L, balance sheet, and cash flow forward using three inputs: actuals (what already happened), drivers (the variables shaping the outcome), and assumptions (what you expect those drivers to do). For SME CFOs, it's the line that separates reporting on the past from steering the business. Without a forecast, your role stays backward-looking. With one, decisions get a model to land in.
In practice there isn't one forecast — there are four, running side by side, each with its own cadence, audience, and job-to-be-done. The rest of this article walks through those four types, plus the three dimensions that qualify every forecast model: method (how you build each P&L line — driver-based, funnel-based, customer-based, and so on), level of detail (consolidated, per entity, per cost center), and technique (what you do with the model — scenarios, sensitivity, what-if).
That four-by-three taxonomy — types × methods × levels × techniques — is how mature FP&A functions structure the work. For an SME finance team, where headcount is tight, the discipline is to be explicit about each layer in every cycle. Which type are you building, with which method, at which level, applying which technique? Confusion in stakeholder conversations almost always traces back to mixing those layers.
The four forecast types: three accrual variants + one cash forecast
The four types aren't synonyms or shades of the same idea. Each has its own role, cadence, and audience. The first three — budget, rolling, latest estimate — are variants on the same accrual mechanics, covering P&L, balance sheet, and cash flow statement at whatever level of detail you choose. The 13-week cash flow forecast plays a different game entirely.
Budget — annual, formal commitment
The budget gets built in Q4 for the year ahead, then locked. It's the formal number you commit to in front of investors, the board, and your management team — the headline you'll be measured against, the anchor for bonus targets, the baseline for variance reporting all year long. Once set, you don't update it. The point isn't operational steering; the point is having a fixed reference to compare reality against.
Level of detail: pick what fits the business — consolidated for simple groups, per entity for multi-entity setups, per cost center where the structure actually matters (clinics, branches, practices). Cadence: once a year.
Rolling forecast — continuous operational steering
A rolling forecast always projects 12 to 18 months out from today. After each month-end close the window shifts: the oldest month drops off, a new month gets added at the far end, and the freshly closed month's actuals flow into the model. Unlike the budget — which is a fixed reference — the rolling forecast is the live picture your team actually steers with. Most teams keep it pragmatic rather than exhaustive; the value is in the cadence.
Investor context. Venture-backed companies tend to lean on the rolling forecast. VC investors want high cadence and an always-current view on a business that's moving fast. Private equity investors typically prefer the half-yearly LE cycle below — the deeper exercise fits how they manage portfolios.
For the full breakdown — horizon choice, driver-based modelling, the four-step monthly cycle, the handoff from the annual budget, variance analysis, and what to look for in tooling — see the complete guide to rolling forecast for CFOs.
Latest estimate (LE) — deep half-yearly rebuild
The latest estimate is a thorough rebuild of the budget for the remaining 3 to 6 months of the fiscal year. Typically run mid-year after the H1 close, sometimes again in Q3. It's not a continuous projection: it's a one-off, deep revision — H1 is in the books, what does H2 honestly look like now?
Strictly speaking, if your rolling forecast is well-maintained, the LE adds nothing the rolling doesn't already give you. But PE investors usually want one anyway, because the LE cycle forces a level of rigour that monthly rolling updates rarely match: every driver reconsidered from scratch, every assumption made explicit, cross-functional input from sales, ops, and HR. Rolling forecasts get short on time; the LE is the deliberate, slow round.
Investor context. Private equity portcos typically face a half-yearly LE alongside the budget and rolling forecast. Venture capital investors usually skip it — they get more value from the high cadence of rolling updates. The difference comes down to time horizon and portfolio style: PE manages mature businesses on full-year cycles; VC manages high-velocity ones on shorter ones.
SMEs without external investors can stop at the rolling forecast. PE portcos and organisations with formal half-yearly reporting keep the LE as a separate cycle — the discipline alone is worth it.
13-week cash flow forecast — short-term liquidity
The 13-week cash flow forecast plays a different game from the other three. Three things set it apart:
Cadence: weekly by default, daily when cash is tight. The accrual forecasts run monthly or quarterly.
Basis: actual cash movements. Open receivables with expected collection dates, open payables with due dates, recurring monthly outflows. No accrual layer in between.
Scope: cash flow only. No P&L, no balance sheet.
It's built for one job: short-term liquidity management. The questions it answers go from “will we make payroll next month?” up to “does this acquisition fit within the next 13 weeks of available cash?”. For the full mechanics — the weekly cycle, AR/AP aging as the foundation, covenant tracking for PE portcos, runway monitoring for scaleups — see the complete guide to the 13-week cash flow forecast.
Together, the four types cover different angles of the same business. Budget gives you commitment. Rolling forecast gives you the live picture. The 13-week cash flow warns you early when liquidity is about to get awkward. The LE is the deliberate deep-dive that PE investors specifically value — without that external pull, the rolling forecast covers the same ground.
A practical rule of thumb: VC-backed businesses run on rolling forecasts; PE-backed businesses run rolling plus LE. Without external investors, your own governance and stakeholder cadence determine which combination is worth the time.
The three qualifying dimensions: method, level of detail, technique
Type is the first axis. Three more cut across all of it: method, level of detail, and technique. They apply to the three accrual types — budget, rolling, LE. The 13-week cash flow has its own internal logic and largely sits outside this grid.
The point of the distinction: too many forecasting conversations conflate these layers. “We do scenario planning” isn't a forecast type — it's a technique that runs on top of any of the four. “We use a rolling forecast” is a type choice; you can then model it driver-based or customer-based, at group level or per entity, with or without scenarios layered on. Naming the layer you're talking about cuts through most of the confusion in stakeholder meetings.
Deeper on each layer: driver-based forecasting, scenario planning, and multi-entity forecasting.
Why forecasting belongs in a spreadsheet
For SMEs, the forecast model belongs in a spreadsheet. Not because tools are bad — because forecast models are too varied to fit any fixed structure.
Revenue alone can be modelled five different ways: driver-based (volume × price), funnel-based (pipeline conversion), customer-based (per-customer ARR), cohort-based (signup-month dynamics), contract-based (per contract with expirations). Each of those can be split by product, geography, entity, cost center, sometimes per person. Cost structure, working capital, and capex projections show the same range. No tool can pick one shape that fits every business.
Any tool that bakes in a structure will fail somewhere — it won't fit one of your product lines, or one of your entities, or the way you actually think about cost-of-sales. Any tool that supports everything ends up being a slower spreadsheet with worse formulas. You'll route around it. You'll keep a real model in Excel anyway, and now you maintain two versions that drift.
Forecasting software earns its keep in two places: large enterprises with dedicated FP&A teams that can absorb months-long implementations, and businesses with genuinely complex supply chains where inventory modelling stops fitting in cells. For an SME finance team, Excel and Google Sheets are the right tool — flexible, familiar, no learning curve, and exactly suited to the architecture variety you hit every cycle.
So the real question isn't “spreadsheet or tool?” It's “spreadsheet plus which tool?”
Consolidation and forecasting in one environment
A forecast is only as good as the actuals underneath it. For multi-entity SME groups that's where most of the friction lives. Consolidating actuals manually in a spreadsheet means version control problems, elimination errors, and a sluggish handoff between “numbers from the ERP” and “numbers in the model”. Without clean consolidated actuals, your forecast has no honest baseline.
What Finstack does differently is closing that loop in one environment, both sides at transaction level.
- Finstack consolidates actuals with three elimination methods (full GLA, per IC relationship, per transaction), intercompany reconciliation at transaction level, parallel reporting for management/statutory/per-BU views, and automatic currency conversion. The full breakdown is in the consolidation pillar.
- Finstack pushes that consolidated data straight into your forecast spreadsheet through native two-way Excel and Google Sheets integration. Actuals refresh into a dedicated [Finstack] Actuals tab; your model recalculates; you push the result back to Finstack with one click.
The alternatives don't close the loop the same way. Speedbooks and Visionplanner are built for year-end compilation, not real forecasting — they work at trial-balance level. Lucanet and BrightAnalytics handle forecasting, but inside their own UI and without the spreadsheet flexibility your modelling actually needs — plus they come with consultancy-heavy implementations in a different price bracket. Specialist FP&A tools usually skip transaction-level consolidation entirely. In the SME segment, Finstack is the only tool we know of that does both, treats the spreadsheet as the modelling surface, and goes live in a day.
How forecasting in Finstack works
Finstack organises forecasting around Forecasts that can hold one or more Scenarios. For each scenario you pick four things:
- Structure — the P&L and balance sheet layout the scenario follows (a Group structure, a per-division structure, whatever fits)
- Method — Manual (entered in the Finstack UI), Google Sheet, or Excel
- Detail level — Consolidated or Per entity
- Default time frame — the standard calendar for the forecast
You enter P&L and balance sheet directly; the cash flow statement derives automatically from both, using the CFS mapping you set in Setup. That keeps the three statements consistent and saves you from rebuilding cash flow logic by hand.
With the Google Sheet or Excel methods, Finstack adds two tabs to your spreadsheet: [Finstack] Actuals (read-only, refreshed automatically — don't edit it) and [Finstack] Forecast (your output, which writes back to Finstack with one click). Your modelling work — drivers, calculations, scenarios — lives in your own tabs (typically a Calculations tab) referencing the two Finstack tabs. No limit on how complex you go.
For the full feature documentation: see the Finstack Help Center page on Forecast.
Start with the one forecast you currently miss most. For most SME CFOs that's either the 13-week cash flow (concrete, short, immediately useful) or the rolling forecast (operational steering). Build it in a spreadsheet using Finstack actuals via the Google Sheet or Excel integration, then expand to budget and LE once the first cycle is running cleanly.
Explore Finstack.
Free trial for 14 days.
Access to all features. No credit card required.
.png)
Three common mistakes in SME forecasting
Trying to model in a tool UI instead of a spreadsheet
Tools that lock down the forecast architecture are structurally too rigid for the variety of models an SME actually runs. You lose depth in stakeholder conversations and end up keeping a real model in Excel on the side anyway. Do the modelling in Excel or Google Sheets. Use a tool like Finstack for consolidated actuals and the forecast roll-up — not for the modelling itself.
Mixing up type, method, and technique
“We do scenario planning” isn't a forecast type — it's a technique that layers on top of any of the four. “We run a rolling forecast” is a type choice; you can build it driver-based or customer-based, at group level or per entity, with or without scenarios. Being explicit about which layer you're discussing cuts the noise in stakeholder meetings and makes governance discussions actually useful.
Treating the 13-week cash flow as “rolling forecast for cash”
It isn't. Weekly, not monthly. Cash transactions, not accrual. Cash flow only, not P&L and balance sheet. Mash them together and you get a cash flow view too coarse for liquidity management and a 13-week cycle too fiddly for operational rolling-forecast cadence. Keep them on separate tracks.
Frequently asked questions
Can't find your question? Let us know
What is forecasting for SME CFOs?
Forecasting projects your P&L, balance sheet, and cash flow forward using actuals, drivers, and assumptions. For SME CFOs it splits into four distinct types that live side by side: an annual budget (the commitment you make to investors and your management team), a rolling forecast (your operational steering instrument, refreshed monthly or quarterly and looking 12-18 months ahead — favoured by venture capital investors), a latest estimate (a deep, half-yearly rebuild of the budget for the remaining months — favoured by private equity), and a 13-week cash flow forecast (weekly, cash-based, focused on liquidity). The first three are accrual variants and can be built at group level, per entity, or per cost center. The 13-week cash flow runs on a different track entirely.
What is the difference between budget, rolling forecast, and latest estimate?
All three are accrual-based and cover P&L, balance sheet, and cash flow — but they play different roles. The budget is set once in Q4, then locked: it's your formal commitment to investors and your board, and it anchors bonus targets all year. The rolling forecast is continuous — you update it monthly after the close, always projecting 12 to 18 months ahead, and it's the operational instrument your team actually steers with. Venture-backed companies tend to lean on this one. The latest estimate is a deeper, half-yearly rebuild: H1 actuals plus a thorough rework of the assumptions for H2. Strictly speaking, if your rolling forecast is good, the LE adds nothing — but private equity investors usually want it anyway, because the LE forces a level of cross-functional rigour that monthly rolling updates often skip.
Why is the 13-week cash flow forecast fundamentally different?
Three reasons. First, cadence: weekly updates by default, daily when cash is tight — while the other three operate on monthly or quarterly cycles. Second, basis: it runs on actual cash movements (open receivables with expected collection dates, open payables with due dates, recurring monthly costs) rather than on accrual accounting. Third, scope: it's a cash flow view only — no P&L, no balance sheet. It's built for a specific job: answering “will we make payroll next month?” or “can we afford this investment in the next 13 weeks?”. That's a different question from the strategic plan-vs-actuals work the other three forecasts support.
Why does forecasting belong in a spreadsheet rather than a tool UI?
Because SME forecast models are too varied to fit into any fixed structure. Revenue alone can be modelled driver-based, funnel-based, customer-based, cohort-based, or contract-based — and each method gets sliced further by product, geography, entity, cost center, sometimes per person. Any tool that bakes in one structure will fail you somewhere; any tool that supports every structure is effectively a worse spreadsheet. Forecasting software earns its keep in two places: large enterprises with dedicated FP&A teams and budgets for months-long rollouts, and businesses with genuinely complex supply chains and inventory dynamics. For everyone else — most SMEs — Excel and Google Sheets are the right tool. Finstack feeds those sheets with consolidated actuals, so you keep the flexibility and lose the manual data plumbing.
What is the benefit of consolidation and forecasting in the same tool?
Your forecast is only as good as the actuals underneath it. Finstack consolidates at transaction level (three elimination methods, intercompany reconciliation, parallel reporting) and pushes the consolidated result straight into your forecast model — at group level and per entity. The handoff between actuals and forecast stops being manual: actuals refresh automatically in your spreadsheet, the model recalculates, variance analysis is there immediately. The alternatives don't close the loop: Speedbooks and Visionplanner aren't built for forecasting, Lucanet and BrightAnalytics force you into their UI without spreadsheet flexibility, and specialist FP&A tools usually skip transaction-level consolidation. In the SME segment, Finstack is the only tool that does both well without a months-long implementation.
How does forecasting in Finstack work?
Finstack organises forecasting around Forecasts that can hold multiple Scenarios. For each scenario you pick four things: the Structure (which P&L and balance sheet layout it follows), the Method (Manual, Google Sheet, or Excel), the Detail level (Consolidated or Per entity), and the Default time frame. With the Google Sheet or Excel methods, Finstack creates two tabs in your spreadsheet — [Finstack] Actuals (read-only, refreshed automatically) and [Finstack] Forecast (your output, which writes back to Finstack with one click). You enter P&L and balance sheet directly; the cash flow statement derives automatically from both using the CFS mapping. All your modelling work — drivers, calculations, scenarios — lives in your own tabs alongside the Finstack tabs, with no limit on complexity.
How often should I update my forecast?
Budget: once a year in Q4, then leave it alone — it's a reference, not a moving target. Rolling forecast: every month, right after the close. Keep it fast and pragmatic; the value is in the cadence, not the polish. Latest estimate: once or twice a year (mid-year, sometimes Q3) — a deeper exercise than a rolling update, and only worth running if stakeholders (usually PE investors) explicitly want it. 13-week cash flow: weekly minimum, daily when cash gets tight. With Finstack's automatic actuals feed the cost per cycle drops sharply — a rolling forecast that takes four hours by hand takes about thirty minutes — which is what makes the cadence sustainable rather than aspirational.
CFO turned Founder - Finstack
Sources and provenance
- Finstack Help Center — Forecast — forecast functionality, Manual/Google Sheet/Excel methods
- Finstack Help Center — Sources — ERP connections for the actuals feed
- Finstack Help Center — Cash overview — working capital views at transaction level
- finstack.io — Reporting & insights — reporting and forecasting functionality
- finstack.io — Spreadsheet sync — two-way Excel and Google Sheets integration
- finstack.io — Consolidation — consolidation and elimination functionality
- finstack.io — Pricing — pricing plans and trial
Last reviewed: 3 June 2026 · Next review: September 2026
Consolidated reporting and forecasting.
