Direct Answer: Start With a Rolling 13-Week Cash Flow Forecast

The best SMB cash flow forecast template is a rolling 13-week cash flow spreadsheet that shows expected inflows, outflows, opening balances, and closing balances for every week. It should support at least three scenarios—base case, expected case, and downside case—and be easy for an owner or bookkeeper to update without specialist accounting knowledge. A 13-week period is short enough to remain current and specific enough to reveal payroll, tax, supplier, and customer-payment pressure before it becomes urgent. For a business expecting substantial seasonal or contract-driven changes, add a separate 12-month monthly forecast rather than trying to force annual assumptions into the weekly view.

Also worth reading: What is a rolling 13 week forecast template and how do I build one for my business? · How Should an SMB Forecast Cash Runway in 2026? · Which SMB Cash Forecast Metrics Should Small Businesses Track in 2026?

The template should distinguish cash timing from accounting profit. An invoice issued today may appear in revenue this month but not reach the bank for 60 or 90 days, while a supplier invoice due in seven days must be paid from available cash. That distinction is the core purpose of the forecast, particularly when receivables and payables are significant. It should also separate owner drawings, tax payments, loan repayments, capital expenditure, and operating expenses so an owner cannot mistake planned spending for genuinely available money.

For most small and medium-sized businesses, Microsoft Excel or a compatible spreadsheet remains a practical starting point. The template is “best” when it is understandable, editable, and connected to reliable opening bank and transaction data—not when it contains dozens of complex formulas. Businesses with multiple entities, several currencies, complex inventory, or frequent acquisitions may need dedicated forecasting software, but a sophisticated tool is not automatically more accurate. As of 29 September 2026, the relevant standard is still disciplined data entry and frequent review rather than the newest software feature.

What the Template Should Calculate

A useful forecast includes an opening cash balance for each week, expected receipts, expected payments, net movement, and a closing balance. The opening balance in week one should equal the cash available at the forecast starting date, not merely the total shown in the bank’s current account statement. If the business holds both operating and reserve accounts, use an opening position that reflects how much can realistically be spent without disrupting committed obligations. Add an “overdraft or funding line” only when management has approved a specific facility; do not treat hypothetical borrowing as existing cash.

The most important calculated warning signal is the lowest forecast cash balance, not the total annual surplus. A forecast can show positive annual net cash flow while recording a severe shortfall in the week a tax bill is due. The template should therefore flag any week below a minimum operating reserve, such as three months of unavoidable expenditure, or below an explicitly approved cash floor. Owners should define that floor based on payroll, core supplier commitments, debt service, and realistic customer delays rather than applying the same percentage to every business.

Work backwards from major obligations when setting timing. If quarterly estimated tax is £8,000 on 31 January, cash must be protected earlier rather than left in the account until the payment date. Supplier payments due on 20 October should appear in week four of a forecast beginning 29 September 2026, while a customer expected to pay 60 days after invoice receipt should be placed against a defensible due date or probability-weighted collection date. The model should also include a small uncertainty allowance because expected dates are estimates, especially where customers pay late or supplier terms are inconsistent.

The template should preserve history as well as forecasts. Keeping actual receipts and payments beside forecast values allows managers to see whether assumptions are realistic and adjust later weeks instead of rewriting the past. A simple variance column showing forecast cash versus actual cash is often enough. More elaborate metrics can help, but they are secondary to an accurate opening balance, realistic payment dates, and a clear minimum-liquidity threshold.

Why SMBs Need Both Weekly and Monthly Views

The weekly view is the operating tool because it catches timing problems that monthly reporting can conceal. A business may collect several large invoices in the final week of a month, making the month appear comfortable even though mid-month payroll and VAT obligations are not covered. Rolling the weekly model every week keeps assumptions current as invoices are issued, invoices are paid, and non-essential spending is postponed or approved. It also gives a small team a shared view of which commitments must be funded before discretionary spending can proceed.

The monthly view is still necessary for planning annual capacity, seasonal inventory, hiring, tax, and capital expenditure. Twelve months of monthly data is generally more useful than one week extrapolated across a year because businesses change as contracts are won, stock is purchased, debt is repaid, and fixed costs rise. Monthly forecasts also make it easier to communicate the expected effect of a proposed investment, price increase, or new employee to directors and owners. The two views should reconcile: the monthly opening cash should match the prior month’s closing cash, and annual totals should roll up from the monthly model rather than from a separate set of unconnected assumptions.

Many businesses forecast one of these periods but not the other. Weekly-only forecasts can exaggerate short-term precision while missing a seasonal loss, and annual-only forecasts can fail to reveal a temporary liquidity crisis. The 13-week cash flow forecast should therefore feed, rather than compete with, a 12-month monthly plan. If time is severely constrained, start with 13 weeks, then add a simplified 12-month view with four or five major income and cost categories once the near-term model is working.

Forecast accuracy should be measured rather than assumed. Accountancy Age reported in its research context that 57% of UK SMEs were warning about rising costs, illustrating why fixed obligations and timing need active attention. Businesses can measure performance by comparing forecast closing cash with actual closing cash, expected payment dates with bank receipts, and planned operating costs with actual spending over at least four to six weeks. If actuals routinely differ by more than roughly 10–15%, investigate the cause before simply changing the forecast to match the outcome.

How to Build the Template in Four Practical Stages

First, establish a reliable starting position by reconciling bank, credit-card, loan, and petty-cash balances to the forecast date. Include amounts that are legally available, expected inflows for that week, and obligations already approved, even if invoices have not yet been issued. Exclude uncertain sales and unsupported asset-sale proceeds, while separately showing them as potential upside if they have reasonable evidence. Using a 29 September 2026 start date, for example, means week one begins with 29 September cash and ends on 5 October; the exact calendar boundaries should be documented.

Second, enter income and expenses using payment dates rather than invoice dates alone. Receivables should carry the customer, invoice amount, due date, current status, expected payment date, and probability of collection. Payables should carry the supplier, amount, contractual due date, payment run, and whether the expense is essential or postponable. Payroll, tax, rent, insurance, debt service, utilities, owner drawings, and planned capital expenditure should each have their own rows because combining them hides the size and timing of commitments.

Third, create at least three scenarios. The base case should use the most likely timing, while the downside case might delay the two largest receivables by 30 days, reduce expected sales by 10–20%, and bring forward a supplier payment. The upside case can show earlier receipts or a confirmed contract without pretending that all pipeline opportunities will materialize. Scenarios should be documented assumptions, not emotional judgments; a downside that reduces sales by 80% may be dramatic but has little decision value unless it represents a realistic stress event.

Fourth, review and refresh the model on a fixed cadence. A rolling 13-week forecast is usually updated weekly, with a more detailed review at month-end when actual results become available. At each review, record actual receipts and payments, replace old expected dates, and ask whether the cash floor and planned actions still apply. If the minimum balance breaches an approved threshold, management should have pre-agreed responses such as accelerating collections, delaying non-essential purchases, drawing on a confirmed facility, or reducing discretionary expenditure.

Spreadsheet, Accounting Software, or AI-Assisted Forecasting?

Spreadsheets are flexible, inexpensive, and familiar, but manual formulas can break when users copy cells, rename categories, or overwrite assumptions. Dedicated accounting and planning software can automate bank feeds, integrate ledgers, and provide scenario controls, yet it may require subscription fees and a steeper learning process. AI can help classify transactions, explain variance, or draft questions about a forecast, but it should not invent customer payment dates or independently decide how much cash is safe to spend. Any AI-produced explanation should be checked against the source data and approved by a person who understands the business.

FeatureSpreadsheet templateAccounting or planning softwareAI-assisted cash coach
Typical costOften £0, with optional file purchaseSubscription or licence varies by productMay be included in software or offered as a separate plan
Setup timeSeveral hours for a basic modelOften days, including configurationRequires a reliable spreadsheet or software base
Data updatesManual unless linked to exportsOften automated after integrationMay suggest updates, but human verification is required
Best useStraightforward weekly cash controlMulti-entity or frequently changing forecastsExplaining variance and coaching cash decisions
Main weaknessFormula and version-control errorsCost, complexity, and data-mapping riskIncorrect assumptions or overconfident recommendations
The choice should follow complexity rather than fashion. A small business with one bank account, stable revenue, and modest variable costs can usually manage with a carefully designed spreadsheet. A business with several entities, 15,000 monthly transactions, multi-currency operations, or frequent scenario changes may gain more from integrated software. An AI-transparent cash flow and savings coach is most useful when it displays its inputs, assumptions, and uncertainty rather than presenting a single unexplained number as certainty.

A hybrid approach often provides the best balance. Keep the authoritative transaction and opening-balance records in the accounting system, produce an editable forecast file, and use an AI tool only for tasks such as categorizing narrative, identifying unusual variance, or explaining which assumption changed. This reduces manual entry without surrendering control. The forecast should never recommend a transfer, payment, or borrowing decision solely from an AI score; the owner remains responsible for legal, tax, and financial judgment.

Common Mistakes That Make the Forecast Misleading

The most damaging mistake is beginning with an incomplete bank balance. Include every operating account and short-term obligation, then decide separately which cash is genuinely available for forecast spending. Another frequent error is booking sales when an order is received rather than when payment is expected, especially in businesses with 30-, 60-, or 90-day terms. This overstates liquidity and can lead to tax, payroll, or supplier-payment problems. Historical payment behavior is more useful than generic optimism.

Forecasts also become unreliable when expected receipts and unconfirmed orders are treated identically. Pipeline sales may be useful as an upside scenario, but they should not support committed expenditure unless the likelihood is high and the timing is documented. A similar problem occurs when monthly averages conceal customer concentration. A forecast showing £50,000 of receipts may still be weak if one customer represents £30,000 and pays 45 days late. Concentration, due dates, and payment history deserve explicit review.

Another mistake is allowing owners to use the same model for management reporting and cash availability. Profit measures exclude some accruals and include others, whereas the forecast must track actual movement in liquid funds. The model should also prevent silent changes to historical actuals. Keep forecast and actual columns separate, lock formulas where practical, and maintain a short change log for material revisions. Otherwise, a team can erase evidence that the original forecast was unrealistic and lose the ability to improve assumptions.

Finally, do not confuse a surplus with a buffer. A business can close the month with positive net cash flow while still being unable to pay a large invoice on the following month’s due date. Set a minimum cash threshold based on unavoidable spending and test it against delayed receipts and rising costs. Accountancy Age’s reported 57% warning among UK SMEs about rising costs reinforces the need for a downside case, but the percentage does not prove that every business will face the same pressure. Local costs, contracts, debt terms, and payment behavior still determine the correct threshold.

When to Act and What Thresholds to Use

The forecast should be prepared immediately if payroll depends on a narrow cash window, customers pay late, supplier terms are short, or tax and debt payments are material. In those circumstances, update it weekly and review the next 13 weeks every Monday or Friday. Businesses with steadier cash flows can update monthly, although a weekly review remains inexpensive and can catch problems earlier. As a general rule, a business with less than three months of unavoidable expenditure available should move from monitoring to corrective planning, but it should not panic-sell long-term assets or default on an obligation without first seeking professional advice.

Useful thresholds include a minimum cash balance equal to three months of unavoidable outgoings, a collection target of five days before a payment becomes due, and an overdue receivable review at seven days past the due date. A business with customer terms of 30 days should not automatically use a 60-day receipt assumption because competitors may offer different terms, but it should measure how often customers actually pay late. A 10% sales reduction or a 30-day delay for the largest receivables can form a practical downside scenario, with harsher tests reserved for genuine risk events.

Review actions at predetermined points. By five working days before payroll, confirm funding for the full gross payroll and related deductions. Before approving a non-essential purchase, test whether it breaches the cash floor or requires an owner decision. At least 90 days before a known tax payment, set aside cash in a separate protected balance if feasible. Where a critical invoice is due within 30 days and no cash is available, contact the supplier or lender early, because early communication is generally more useful than a missed payment.

The timing of action matters because waiting until the forecast is negative reduces available choices. If the downside case falls below the cash floor for more than one week, begin contingency planning before that date. If the base case breaches the floor, verify the assumptions first, then consider collection acceleration, payment rescheduling, inventory reduction, and confirmed financing. Avoid relying on unapproved credit, and avoid presenting optimistic collections to lenders or staff as guaranteed funding.

Pricing, Implementation, and the Definition of Useful

A basic SMB cash flow forecast template can be free, with costs limited to the time required to design, populate, and maintain it. Paid spreadsheet templates may be useful if they include clear instructions, data validation, and scenario controls, but price alone does not establish quality. Business Insider’s guide to Excel for small businesses, cited in the research context, supports the broader point that a capable spreadsheet can help a small firm go beyond basic bookkeeping and reduce unnecessary software costs. That does not mean spreadsheet modelling replaces an accountant or that free tools eliminate the cost of staff time.

Implementation should be staged. A one-week cash view with opening balance, receipts, payments, and closing balance can be built in a few hours by a competent bookkeeper. A dependable 13-week model with receivables, payables, payroll, tax, reserve thresholds, scenarios, and variance tracking may take several days initially, especially if historical data must be cleaned. Reviewing and updating it weekly takes less time once category definitions and source systems are stable, but the exact duration cannot be guaranteed without knowing transaction volume and accounting quality.

The template delivers value only if it changes a decision. Owners should record actions such as “invoice customer X two days earlier,” “postpone equipment purchase by four weeks,” or “maintain £12,000 as minimum operating cash.” Track whether these actions improved actual liquidity and whether forecast accuracy improved. If the model is never used to approve spending, sequence collections, or discuss funding, it is a report rather than a cash flow management system.

For glassjar.co, the appropriate editorial position is practical and measured: transparent AI can help an SMB understand assumptions, compare scenarios, and notice cash shortfalls, while the owner or accountant remains the decision-maker. The product angle is therefore not that AI predicts the future, but that it makes the forecast’s inputs and trade-offs easier to inspect. On 29 September 2026, that remains a more defensible promise than claiming software can remove uncertainty or guarantee savings.