🧭 Overview
This guide explains how to forecast cash flow using a practical, repeatable workflow built on FreeAgent exports and Model Reef. It’s for founders, finance leads, and advisors who need a forecast they can refresh weekly without rebuilding spreadsheets. You’ll learn what data to extract from FreeAgent, how to map it into a trusted structure, and how to turn it into forward-looking cash timing with drivers and scenarios. The output is a decision-ready weekly view you can use to protect runway, plan spend, and communicate clearly with stakeholders. For the wider ecosystem and context, see FreeAgent cash flow forecasting.
🤝 How Model Reef + FreeAgent Fit Together
FreeAgent remains responsible for accounting truth: reconciled transactions, outstanding invoices and bills, and historical reporting. It answers “what happened” and supports compliance and auditability. Model Reef is responsible for planning truth: drivers, timing, scenarios, and consistent outputs that answer “what happens next.” That’s the difference between reporting and cash flow forecasting.
Data moves one way: FreeAgent exports (actuals) feed Model Reef (planning). Your forecast logic stays in Model Reef so it doesn’t get rewritten every cycle. You can then refresh actuals on a schedule and update assumptions intentionally-without breaking links or losing version control. This is especially valuable when cash timing is sensitive (payroll, tax, large receivables) and you need a single model that can produce a weekly plan and scenario views from the same structure. If you want a faster starting point, begin with Templates. This pairing is best when you want reliable refresh cycles, scenario agility, and a forecast you can explain.
Responsibilities & Hand-Offs (required)
| Category |
FreeAgent |
Model Reef |
| Source-of-truth system |
Holds reconciled accounting actuals. |
Holds the forecast model and scenario logic. |
| Primary job-to-be-done |
Track and report what already happened. |
Plan and stress-test what happens next. |
| Data captured / managed |
Invoices, bills, bank transactions, journals. |
Drivers, timing assumptions, scenarios, outputs. |
| Data exported / shared |
Actuals exports and reporting summaries. |
Forecast packs, scenario comparisons, runway views. |
| What gets modeled in Model Reef |
Not built for forward modelling structure. |
The workflow for how to forecast cash flow reliably. |
| Refresh cadence |
Continuous as transactions are posted. |
Weekly/monthly refresh with controlled validation steps. |
| Ownership |
Owned by the ledger/accounting owner. |
Owned by finance who maintains assumptions and logic. |
| Outputs produced |
Historical statements and reports. |
cash flow forecast example outputs and scenario views. |
| Common failure point |
Incomplete coding or late posting distorts data. |
Timing assumptions not maintained causes drift. |
| Best-practice guardrail |
Enforce consistent categorisation discipline. |
Map once, refresh consistently, track variances weekly. |
✅ Before You Begin
Before you follow the steps for how to forecast cash flow, set these foundations:
- Access/permissions: confirm you can export FreeAgent data and verify balances (especially bank accounts).
- Data needed: cash/bank balances, receivables (outstanding invoices), payables (outstanding bills), payroll schedule, tax schedule, and core operating expenses.
- Mapping decisions: define the categories that matter for decisions (collections, payroll, rent, tax, overhead, variable spend).
- Refresh cadence decision: choose weekly as the default, with clear triggers for ad-hoc refresh (payroll run, major receivable, large supplier payment).
- Ownership decision: assign one person to maintain mapping and another to own scenario assumptions (with a backup for continuity).
- Assumption boundaries: decide what’s a driver vs what’s a one-off override, so your cash forecasting remains explainable.
If you want to simplify the import/refresh pathway over time, start with Integrations to pick the cleanest long-term workflow. You’re ready if your FreeAgent exports are consistent and you can commit to a weekly review loop.
Step-by-Step Instructions
Step 1: 🎯 Define the workflow and success criteria.
The first step in how to forecast cash flow is agreeing what the forecast must achieve. Define your horizon (commonly 13 weeks for operational cash, or rolling weekly/monthly for longer planning). Define the audience: founder decisions, finance operations, board updates, or client advisory. Then define the minimum outputs: opening cash, receipts, payments, net movement, closing cash, and a minimum cash threshold. Add one “explainability” requirement: your team must be able to answer why cash changes week-to-week (timing, volume, or one-offs). Finally, define success metrics: forecast accuracy by week, time-to-refresh, and number of manual overrides. These metrics keep your process improving rather than drifting into spreadsheet habit.
Step 2: 🔌 Extract/connect the data cleanly.
Export a consistent FreeAgent dataset on a fixed schedule. Before you import anything, run sanity checks: bank balances should match expected closing cash, and receivables/payables should be directionally aligned with what the business believes is outstanding. Import into Model Reef using consistent file naming and stable reporting windows so refreshes don’t introduce noise. If you want fewer manual steps and more reliable refresh cycles, align your workflow to Deep Integrations so mapping and updates can remain persistent over time. The goal is repeatability: the data step should take minutes, not hours. Clean inputs are what make cash flow forecasting feel controlled rather than reactive.
Step 3: 🧩 Map and reconcile (lock the source of truth).
Mapping is where most “how to do a cash flow forecast” guides stay vague-but it’s the part that determines whether you can trust the result. In Model Reef, map cash accounts first, then map revenue and expenses into a small set of planning categories. Reconcile totals against FreeAgent reports so you know the imported baseline matches reality. Then add timing rules: expected collection timing for invoices and expected payment timing for bills. This is the bridge between accounting and cash reality. If categories or labels drift over time, update the mapping once and keep it governed-don’t patch outputs manually. When mapping is clean, you can refresh actuals and the forecast remains stable.
Step 4: 🏗️ Build the model logic + outputs.
Now convert structure into a working forecast. Build receipts schedules from assumptions: collection timing, seasonality, and any known large customer events. Build payment schedules from payroll cadence, tax timing, recurring subscriptions, and supplier terms. Then generate outputs that read like a weekly story: opening cash → receipts → payments → closing cash. Add scenario levers that match real decisions: delay collections by 14 days, reduce discretionary spend by 10%, or shift supplier payment timing. This creates a repeatable cash flow forecast example you can update without rebuilding. The best outputs are simple, consistent, and decision-ready-so leadership can act quickly when cash risk shows up.
Step 5: 🔁 Operationalise: cadence + governance.
The final step in how to forecast cash flow is turning it into an operating rhythm. Refresh weekly, validate, review variance, and publish outputs. Assign ownership: one person maintains mapping and model structure, another updates assumptions, and stakeholders review results. Document a short refresh checklist so the process survives team changes. Track accuracy over time (last week’s forecast vs this week’s reality) and improve one driver per cycle. This is how forecasting matures: fewer overrides, clearer assumptions, better scenario discipline. When governance is consistent, cash flow forecasting becomes a management system-not a spreadsheet emergency. The payoff is confidence: the model stays explainable, refreshable, and aligned to the decisions the business must make.
🧠 Tips, Edge Cases & Gotchas
- If you’re learning how to forecast cash flow for the first time, start with a weekly horizon and only 6–10 categories. Complexity can come later.
- Separate timing assumptions from volume assumptions. If you mix them, it becomes hard to explain why cash moved.
- Keep one baseline scenario “clean” and put experimentation into named scenarios; otherwise your forecast loses credibility.
- Don’t let one-off events distort the baseline. Isolate large tax adjustments, equipment purchases, or exceptional refunds.
- If collections are lumpy, segment receivables by customer type or payment behaviour so your timing rules are more realistic.
- If you want a comparative workflow outside FreeAgent, FreshBooks cash flow forecast can help you see which steps stay consistent across accounting platforms and which steps are tool-specific.
📌 Example
A SaaS consultancy wants to improve predictability. Each Friday, they refresh FreeAgent exports and update Model Reef assumptions. Their receipts are driven by customer payment behaviour (most pay within 10-20 days), while payments are scheduled by payroll cadence, tax set-asides, and recurring vendor subscriptions. They run a downside scenario where collections slip by 14 days and churn increases, reducing receipts. The model shows that cash dips below the minimum buffer in week 7-so they respond early: pause discretionary spend and adjust contractor hours. The following week, actual collections land, and the refresh updates the forecast without rebuilding. This is a practical cash flow forecast example: a simple loop that turns actuals into weekly decisions. To see the workflow in a product walkthrough, use See it in action.
❓ FAQs
Start with a 13-week weekly view, keep categories minimal, and focus on timing. Export bank balances, invoices, and bills from FreeAgent, then build schedules for receipts and payments based on realistic assumptions. Avoid overengineering the model at the start-accuracy improves as you review variance and refine drivers. The key is consistency: refresh weekly, validate totals, and adjust assumptions intentionally. If you’re unsure where to begin, start with a baseline and one downside scenario tied to your biggest cash risk. You’ll build confidence quickly once the workflow becomes routine.
The workflow is similar, but the drivers differ. Services businesses often focus on collections timing, payroll cadence, and utilisation-driven expenses. Product businesses add inventory timing, supplier terms, and working-capital swings. In both cases, the forecast succeeds when timing is explicit and assumptions are documented. Start with a shared structure (opening cash, receipts, payments, closing cash), then layer business-specific drivers. If you keep mapping and governance consistent, your forecast remains explainable even as the business changes.
Accuracy comes from conservative assumptions and continuous review. Use historical collection behaviour (not desired behaviour), include a buffer for timing slippage, and isolate one-offs rather than hiding them in recurring lines. Run at least one downside scenario each week, and track forecast vs actual variance so you can improve the drivers. Also, avoid mixing accounting recognition with cash timing-this is where forecasts become misleading. A realistic forecast is one you can defend and act on, not one that simply “looks good.”
A good cash flow forecast example includes a clear horizon, a small set of categories, explicit timing assumptions, and a variance review loop. It should show opening cash, receipts, payments, and closing cash by week, with a minimum cash threshold and at least one scenario that reflects a real risk. It should also be refreshable: when actuals update, the forecast updates without rewriting formulas. If your example leads to decisions-spend controls, hiring timing, payment renegotiation-it’s doing its job. Start simple, then refine drivers as you learn.
🚀 Next Steps
You now have a clear, repeatable approach for how to forecast cash flow using FreeAgent exports and Model Reef built for weekly refresh cycles, explainable assumptions, and scenario-ready outputs. Next, run the workflow for two consecutive weeks and track variance: that’s the fastest way to improve accuracy and build trust internally. Once the baseline is stable, upgrade one “manual” line into a driver each cycle (collections timing, payroll cadence, or variable spend), and you’ll quickly move from reactive updates to proactive cash control. If you’re rolling this out across teams or clients, standardise mapping and ownership so the model stays dependable as you scale.
Back