Cash Flow Forecasting Software: A Practical Guide

A forecasting tool takes three inputs — your current bank balance, your committed future movements, and your assumptions about everything not yet committed — and produces a projected balance for each future period. The value is not in the arithmetic, which a spreadsheet can do. It is in the fact that the committed movements are pulled automatically from your accounting system and refreshed every day, so the forecast is never more than 24 hours stale.

The five jobs a good system handles

• Sync the ledger — open invoices with due dates, unpaid bills, recurring transactions and bank balances from QuickBooks, Xero, Sage, FreeAgent or a direct bank feed.
• Model payment behaviour — apply the real average days-to-pay per customer rather than the terms printed on the invoice.
• Layer manual assumptions — a hire in March, a tax bill in July, a piece of equipment in Q4, a loan repayment schedule.
• Run scenarios side by side — best case, expected, and the one where your largest customer leaves.
• Report clearly — a projected closing balance per week or month, the date of the lowest point, and the size of any gap.

Direct versus indirect forecasting — and why small businesses use direct

An indirect forecast starts from projected profit and adjusts for non-cash items and working-capital movements. It is what lenders and boards often expect, and it works over quarters and years. A direct forecast lists actual expected receipts and payments, week by week. It is far more accurate in the short term and far easier to explain to a non-accountant.

Almost every small business should run a direct thirteen-week forecast as its operating tool, and an indirect annual forecast only when a bank or investor asks. Thirteen weeks is the standard horizon because it is roughly one quarter — long enough to act, short enough that your assumptions still hold.

1. You check the bank balance before deciding whether to approve a purchase.
2. You have delayed a supplier payment in the last six months to cover payroll.
3. You cannot say what your cash position will be in eight weeks without building a spreadsheet.
4. Your VAT, GST or corporation tax bill has ever been a surprise.
5. Revenue is growing and cash is getting tighter — the classic signature of working-capital strain.
6. You are considering a hire, a lease or a stock purchase and are deciding by feel.
7. A single customer represents more than 20% of revenue and you have never modelled losing them.

Three or more is the threshold at which the discipline pays for itself. Tools in this category run from roughly $20 to $150 a month; one avoided overdraft fee or one better-timed stock order usually covers a year.

The tools below are the ones most commonly recommended by accountants serving small businesses across the US, UK, Canada and Australia. Prices are entry-tier from public pricing pages at time of writing and are typically billed per business rather than per user; always confirm before buying, and check whether your accountant already holds a partner licence that covers you.

Float — the default recommendation for a reason

Float's model is a rolling cash view built directly on your Xero, QuickBooks Online or FreeAgent data, where every invoice and bill appears as a movable item you can drag to the week you actually expect it. That drag-and-drop honesty is the feature: it forces you to record what you believe rather than what the invoice terms claim. Scenario planning is unlimited, the visualisation is genuinely clear in a board meeting, and the sync is reliable. It is also the most expensive of the mainstream options, and it expects your bookkeeping to be current — garbage in, confidently graphed garbage out.

Fathom — when you want reporting and forecasting together

Fathom started as a management-reporting tool and added forecasting, so you get KPI tracking, three-way forecasting (profit, balance sheet and cash) and presentation-ready reports in one subscription. For a business that reports to a board, a lender or an external investor, that combination is efficient. The forecasting workflow is more structured and less tactile than Float's, which suits monthly planning better than weekly cash management.

Futrli — fastest to a first forecast

Futrli leans on prediction: it reads your ledger history and generates a forward view with minimal configuration, which is genuinely useful if the alternative is having no forecast at all. Treat the automatic numbers as a starting draft rather than a plan, and override anything the algorithm cannot know — a contract ending, a seasonal dip, a planned hire.

Agicap — for multi-entity and multi-currency complexity

Once you have two legal entities, several bank accounts and more than one currency, the lightweight tools start to strain. Agicap consolidates across entities with direct bank connections and handles FX properly. Pricing is quote-based and clearly aimed at businesses past the very small stage.

The native tools in QuickBooks and Xero

Both platforms ship short-horizon cash views at no extra cost. Xero's cash-flow view and QuickBooks' cash-flow planner will project roughly 30 to 90 days from existing invoices, bills and recurring transactions. They are limited on scenarios and manual adjustments, but they are free, they are already connected, and they are far better than nothing. Check what your plan includes in the Xero Central documentation or the QuickBooks support centre before paying for a third-party tool you may not need yet.

Start here if: your bookkeeping is current, you have one entity and one currency, and you have never built a forecast before. Upgrade when you need scenarios. If your ledger itself is the weak point, our guide to the best accounting software for freelancers and small firms is the right first step.

The tool does the arithmetic. The accuracy comes from how you set it up. This is the sequence accountants use.

Step 1 — Establish an honest starting balance

Reconcile every bank account to yesterday. A forecast built on an unreconciled balance is wrong before the first week. Include credit-card balances and any overdraft or revolving facility as a separate line, not as available cash.

Step 2 — Load committed inflows with real payment dates

Pull every open sales invoice from your ledger, then adjust the expected date using each customer's actual behaviour. If a client has paid on day 58 for the last six invoices, model day 58 — not the day 30 on your terms. Most tools calculate this average for you; use it.

Step 3 — Load committed outflows, including the ones people forget

• Supplier bills with due dates from the ledger.
• Payroll, including employer taxes and pension or superannuation contributions.
• Sales tax, VAT or GST payments on their statutory dates.
• Corporation or income tax instalments.
• Loan and lease repayments, split between capital and interest.
• Annual software renewals — the ones that quietly hit in the same month every year.

Step 4 — Add uncommitted assumptions separately

Keep pipeline revenue on its own line, clearly labelled, and weight it by your actual win rate rather than optimism. This is where a clean sales pipeline pays a second dividend: the forecast inherits its quality from your CRM. If you are tightening that up, our guide to sales pipeline software covers the stage discipline that makes weighted revenue meaningful.

Step 5 — Review weekly, same day, thirty minutes

Compare last week's forecast to what actually happened, note the variance and its cause, then roll the horizon forward one week. Forecast accuracy improves faster from this half-hour habit than from any amount of extra modelling detail.

Scenario planning is the reason to pay for software rather than keep a spreadsheet. Build these four once and refresh them quarterly.

Reading the results

For each scenario, note three things: the lowest projected balance, the week it occurs, and how many weeks of notice you would have. A gap eleven weeks out is a planning problem with several good solutions — invoice financing, a facility, a payment plan with a supplier, a delayed hire. The same gap discovered with eight days' notice is an emergency with expensive ones.

Levers that move the number fastest

• Invoice on the day of delivery, not at month end. Frequently the single largest improvement available, and it is free.
• Take deposits — 30–50% up front on project work changes the shape of the entire curve.
• Automate chasing. A polite reminder at day 3 before due, day 1 after and day 7 after typically pulls average payment in by a week.
• Offer card and direct-debit payment. The processing fee is almost always cheaper than the financing cost of waiting.
• Negotiate supplier terms in the same conversation as volume. Thirty extra days from a key supplier is working capital you do not have to borrow.

A forecast is only as good as the data feeding it, which means the surrounding systems matter as much as the tool itself.

Accounting ledger — the foundation

Bank feeds must be reconciled at least weekly and ideally daily. If your books close six weeks after month end, no forecasting tool can help you; fix that first, either with better bookkeeping discipline or a bookkeeper.

Invoicing and receivables

Automated reminders and online payment links shorten the collection cycle directly, which is the highest-leverage input into the model. Our roundup of invoicing tools for small business compares the options.

Expense and payables

Approval workflows and card controls stop unplanned spend from appearing in the ledger after the fact — the most common cause of an unexplained forecast variance. See our expense management software guide for how that layer works.

Payroll

Payroll is usually the largest and most predictable outflow, so it should be a scheduled item rather than a manual entry. Most forecasting tools read it from the ledger; verify that employer taxes and pension contributions are included on the correct dates, not lumped into the net pay run.

Banking

Open banking connections give a live balance without waiting for the ledger to catch up. In the UK and EU this is standardised; in the US, aggregation is provided through services such as Plaid. Direct bank feeds are a useful cross-check that the ledger-derived balance is right.

• Forecasting profit instead of cash. Revenue recognised in March that arrives in June does not pay March's wages.
• Using invoice terms instead of payment behaviour. Model what customers do, not what they agreed to.
• Forgetting tax. VAT, GST and sales tax collected on behalf of the government sit in your account temporarily. Treating them as available cash is the single most common cause of a small-business cash crisis.
• Over-modelling. Two hundred line items you cannot maintain will be abandoned within a quarter. Twenty accurate lines beat two hundred stale ones.
• Building it once. A forecast reviewed weekly is a management tool; a forecast built for a bank application is a document.
• Excluding the owner. Dividends, drawings and personal tax are real outflows and belong in the model.
• Ignoring seasonality. Last year's monthly pattern is usually a better guide than a straight-line average.

Done properly, cash flow forecasting changes the character of decision-making in a small business. Hiring, stock purchases, equipment, marketing spend and pricing all stop being nervous judgement calls and become questions with a number attached. That is the whole return on the subscription — not the chart, but the confidence to commit.