A spreadsheet cash forecast costs a finance team four things: the hours spent keeping it current, the errors it carries, the copies it spawns, and the dependence on the person who built it. No independent study puts a figure on any of those for a business of 11 to 50 people, so this page gives you a model to cost your own forecast, with the statutory inputs sourced and every other assumption yours to change.

The four cost lines a subscription page never shows

When a finance team weighs a dedicated forecasting tool against the spreadsheet it already runs, the comparison usually stops at the subscription. We cover that side in what cash flow forecasting software costs. This page is the other side of the same ledger.

The spreadsheet is not free. It is paid for in a currency that never appears on a purchase order: finance-team hours, rework after a broken formula, an afternoon reconciling two versions of the same file, and the week the forecast stops because the one person who understands it is away. Those four lines are the maintenance cost. ICAEW's Twenty Principles for Good Spreadsheet Practice treats each of them as a recognised risk that organisations should control; what the principles do not do, and what nobody has published, is put a price on them. This page builds that cost model, and shows its working.

Cost line 1: the hours

Keeping a cash forecast alive in a spreadsheet is a weekly cycle of four jobs. Someone gathers the current balances from the bank and the ledger. Someone keys in or refreshes the open invoices and unpaid bills. Someone reviews the expected payment dates, because the due date on an invoice is rarely the date the cash arrives. And someone reconciles the model back to the accounts, because if the two disagree the forecast cannot be trusted at the next meeting.

How long that takes depends on transaction volume, how the workbook is built and how often it is refreshed, and no statistical body or professional body measures it. Finance teams that have moved off spreadsheets describe the before and after in their own terms. One fractional CFO on our site puts it this way: "Between downloading bank data, correcting for actuals, changing forecasts, and so on, I'd spend up to two hours a week in Excel." A group financial manager describes the other side: "The setup takes a little effort, but maintenance is less than an hour a week." Those are individual experiences, not benchmarks, and they should not be averaged into one. Your own number belongs in the model below.

Cost line 2: the errors

The research on errors in operational spreadsheets is older than most finance teams would guess. The most thorough field study, by Stephen Powell, Kenneth Baker and Barry Lawson at the Tuck School of Business in 2009, audited fifty workbooks in real use and found formula errors in a small proportion of cells. The finding that matters for a cash forecast is not the percentage but the arithmetic behind it: a low per-cell rate becomes a high per-model rate once a workbook grows to thousands of formulas, links between sheets and weekly edits by more than one person.

Nothing since has replaced that study with a newer measurement of errors in in-use spreadsheets. The headline figures that circulate online trace back to laboratory tasks or to press releases rather than to audits of working files. The spreadsheet failures that reach the news are the ones at banks and government departments, because those are large enough to be reported. There is no register of the small ones. That absence is not evidence that a five-person finance team's forecast is error-free; it is evidence that nobody counts.

What an error costs is specific to what it does. A formula pointing at last month's balance row shows headroom that is not there. A bill keyed with the wrong date moves a cash low by a week. The cost is either the rework to find and fix it, or the decision made on the wrong number before anyone noticed. The model treats it as your assumption for a reason.

Cost line 3: the versions

A forecast spreadsheet shared by several people tends to end up with a rule nobody wrote down: one person edits, everyone else waits. Research into how finance teams actually collaborate in spreadsheets, published by Xia, Sarkar, Brumby and Cox in 2025, observed exactly that pattern. Teams pass a workbook in sequence, keep local copies, and reconcile the differences by hand when the copies drift. The population in that study was broader than a small-business finance team, so treat it as a description of the mechanism rather than a measurement of your team.

The version problem has two costs. The visible one is reconciliation time: the hour spent working out why the forecast on the shared drive and the forecast in Tuesday's email disagree. The less visible one is that a meeting can open on a number produced before the latest payment run. ICAEW's principles ask organisations to keep a consistent backup and version-control process precisely so that there is no ambiguity over which file is current.

Cost line 4: the person

Every spreadsheet forecast has an author, and most finance-team workbooks carry that author's logic in their structure: which sheet feeds which, why a cell is hard-coded, what a colour means, which rows are safe to overwrite. When that person is on leave, off sick or has left, the forecast either stops or someone rebuilds it from the ledger.

ICAEW's Twenty Principles address this directly by asking for documentation that lets any capable user understand and maintain a workbook, for shared ownership, and for peer review. A forecast that meets those principles has low key-person risk. A forecast that does not has a cost that only becomes visible in the week it falls due: the cover hours, the delayed reporting pack, and the decisions taken without a current cash view.

How to cost your own spreadsheet forecast

The only figures in this model that are sourced are the wage and statutory inputs. Everything else is an assumption, labelled as one, for you to replace with your own numbers.

The Office for National Statistics publishes median hourly pay by occupation in the Annual Survey of Hours and Earnings (Table 14, provisional 2025 data released 23 October 2025). For 2026/27, an employer pays Class 1 National Insurance at 15% on earnings above the £5,000 secondary threshold, and a minimum automatic-enrolment pension contribution of 3% on qualifying earnings between £6,240 and £50,270. Loaded cost is wage plus those two on-costs, divided by paid hours. The table uses 37.5 hours a week for 52 weeks (1,950 paid hours) and the minimum pension.

The pension on the financial manager row is capped because pay exceeds the top of the qualifying band. These are national medians across all employers, not small-business figures; if you know your own payroll cost, use it instead.

Step 1: Price an hour of finance-team time. Take the loaded hourly cost for each role that touches the forecast from the table above, or substitute your own payroll figures. Loaded cost is wage plus employer National Insurance and the minimum pension contribution, divided by paid hours.

Step 2: Count the maintenance hours. As an illustration only: a financial accounts manager spends three hours a week on the four jobs in cost line 1 and a financial manager spends one hour reviewing it, for 48 working weeks. That is (3 × £27.79 + 1 × £41.75) × 48, or about £6,000 a year in labour before anything goes wrong. Change the hours and the roles to match your team.

Step 3: Add the rework. Estimate the hours spent each year finding and fixing errors, reconciling versions, and covering for the forecast's owner. Price each at the loaded rate of the person who does it. None of these has a published benchmark; your last twelve months are the best evidence you have.

Step 4: Add the cost of a miss. If a forecast misses a cash low and the business borrows to cover it, the cost is the shortfall multiplied by your facility's annual rate, multiplied by the days it runs, divided by 365, plus any fixed fees. If the miss means paying a supplier late, statutory interest on commercial debts in the UK runs at 8% over the Bank of England base rate; the current base rate is on the Bank's website. We cover the timing side in how payment terms and late payment shape your cash timing.

Step 5: Put the switching cost on the other side. Moving a forecast into a dedicated tool has a one-off cost in set-up hours, and it belongs in the same model rather than left out. Estimate the hours to connect the accounting platform, check the first forecast against the spreadsheet, set up scenarios and show the team how to read it, and price them at the same loaded rates. Then compare the annual maintenance cost from Steps 2 to 4 with the subscription plus the one-off set-up.

Finance teams in the US, Australia and New Zealand can run the same model with their own wage figures. The statutory employer on-costs are 7.65% combined Social Security and Medicare in the US, a 12% Superannuation Guarantee in Australia from 1 July 2025, and a 3.5% minimum employer KiwiSaver contribution in New Zealand from 1 April 2026.

How Float takes out the four cost lines

Float is a cash flow forecasting tool that connects to Xero or QuickBooks Online and builds the forecast from the ledger. Each cost line maps to a specific part of how it works.

The hours in cost line 1 come from gathering, keying and reconciling. Float imports reconciled balances, open invoices and unpaid bills from the accounting platform automatically every 24 hours, at an hour you choose, with a manual sync when you need it sooner. Expected payment dates are taken from the accounting platform where they exist and fall back to the due date where they do not, and they can be moved one at a time or in a batch.

The errors in cost line 2 come from formulas and re-keying. In Float there are no formulas to break and nothing to re-key: the data flows one way, from the accounting platform in, and what is in the forecast is what the ledger holds. That does not make the forecast right about the future, but it removes the class of error that comes from transcription.

The versions in cost line 3 come from copies. Float holds one forecast per company, with four roles: Owner, Admin, Editor and Viewer. Editors change the forecast, Viewers read it, and nobody works from a copy. Comments sit on the invoice or bill they explain.

The person in cost line 4 is the author of the workbook. In Float the structure is the tool's, not an individual's, so a colleague covering the forecast opens the same screens with the same data. Sage Intacct is on our waitlist; it is not live today.

When a spreadsheet is still the right answer

Not every finance team should move. If one person maintains the forecast, transaction volume is low, and the model is reviewed monthly rather than weekly, the four cost lines may add up to less than a subscription. We say so in our guide to the alternatives to spreadsheets for cash flow planning, and the model above is built to give you that answer if it is the true one. The point is to cost the spreadsheet properly before deciding, rather than to treat it as free because nothing is invoiced.

Frequently asked questions

How many hours does a finance team spend maintaining a spreadsheet cash forecast?

No independent study measures the hours a finance team spends maintaining a spreadsheet cash forecast at any company size. The work is a weekly cycle of gathering balances, refreshing invoices and bills, reviewing expected payment dates and reconciling the model to the ledger. Time your own cycle for a month and use that figure; customer descriptions on this site range from under an hour to a few hours a week, and they are individual experiences rather than benchmarks.

What is the error rate in a spreadsheet cash forecast?

There is no reliable error rate for spreadsheet cash forecasts specifically. The best field evidence on operational spreadsheets in general, from a 2009 Tuck School of Business audit of fifty workbooks, found formula errors in a small proportion of cells, which becomes material as a model grows. Widely repeated headline figures come from laboratory studies or press releases rather than audits of working files, and should not be relied on.

Why does a shared forecast spreadsheet end up with several versions?

A shared forecast spreadsheet ends up with several versions because teams pass the file in sequence, keep local copies for their own edits, and email snapshots for meetings. Each copy drifts from the others until someone reconciles them by hand. ICAEW's spreadsheet principles ask for a consistent version-control process so that there is never ambiguity over which file is current.

What is key-person risk in a spreadsheet cash forecast?

Key-person risk in a spreadsheet cash forecast is the dependence on the one person who understands how the workbook is built. When that person is absent or leaves, the forecast stops or has to be rebuilt. ICAEW's Twenty Principles for Good Spreadsheet Practice recommend documentation, shared ownership and peer review so that any capable user can maintain the workbook.

What does an hour of finance-team time cost in the UK?

An hour of finance-team time in the UK costs the median hourly pay for the role plus employer National Insurance at 15% above the £5,000 secondary threshold and a minimum 3% pension contribution on qualifying earnings, spread over paid hours. On the ONS medians, that is roughly £42 an hour for a financial manager, £28 for a financial accounts manager and £19 for a book-keeper or payroll clerk, at 1,950 paid hours a year. Use your own payroll figures where you have them.

What does it cost when a cash forecast misses a low point?

When a cash forecast misses a low point, the cost depends on what happens next. Borrowing to cover the gap costs the shortfall multiplied by your facility's annual rate for the days it runs, plus any fees. Paying a supplier late in the UK exposes you to statutory interest at 8% over the Bank of England base rate, and a missed early-payment discount costs whatever the supplier offered.

How long does it take to move a spreadsheet cash forecast into Float?

Moving a spreadsheet cash forecast into Float starts with connecting Xero or QuickBooks Online, after which the forecast is built from the ledger's reconciled balances, open invoices and unpaid bills. The set-up effort beyond that is checking the first forecast against your spreadsheet, adjusting expected dates and adding scenarios. One customer describes it as "The setup takes a little effort, but maintenance is less than an hour a week"; your own time will depend on how many adjustments your spreadsheet carried.

Who can see and edit the cash forecast in Float?

Access to the cash forecast in Float is set by four roles. The Owner (one per company) and Admins have full edit access and manage users, Editors have full edit access without user management, and Viewers have read-only access. Data flows one way from the accounting platform into Float, so nothing done in Float writes back to Xero or QuickBooks Online.

If you want to see what your forecast looks like built from the ledger rather than a workbook, start a free trial and connect your accounting platform.