Mid-market finance leaders rank cash flow forecasting as the hardest part of the job, ahead of reconciliation and data integration. The companies changing fastest are the ones still leaning hardest on spreadsheets to hold it together.
Key Takeaways
Ask a mid-market CFO what keeps the finance function from scaling and the answer is no longer the close or the audit. It is the cash forecast. New survey data shows that forecasting has become the pressure point where growth, new business lines and fragmented systems all meet, and very few finance teams believe the tools underneath it are ready for what the next two years will bring.
PYMNTS Intelligence's Growth and Scaling: The Corporate Finance Inflection Point, produced with Fynapse, surveyed 60 CFOs and senior finance executives at U.S. companies with annual revenue between $100 million and $1 billion from June 9 to 16, 2026. Sixty-two percent said cash flow forecasting was difficult to manage or scale, and 37% named it their single hardest finance or back-office challenge. No other area came close: data integration, receivables and collections, and reconciliation each landed between 12% and 15% as the top problem.
The difficulty does not depend on growth. As PYMNTS reported on November 4, a company does not need rising sales to lose sight of its cash. Among firms without revenue growth over the past 24 months, 48% added suppliers, 40% handled higher transaction volumes and 56% added a product or service line. Every one of those changes adds payment terms, counterparties and timing assumptions to a forecast that has to absorb them, whether or not the top line moved.
Finance leaders know where the money should go. Cash flow forecasting was the top investment priority at 52%, ahead of real-time reporting at 35%, reducing manual processes at 33% and unifying finance data at 32%. Yet only 12% said their finance and back-office systems were fully prepared for the next two years, even though 62% described them as at least mostly prepared. By sector, 31% of technology firms said they were fully ready, against 10% of services firms and none of the goods companies.
The sharpest split in the data is between companies that are changing on several fronts at once and those that are not. Forty-five percent of respondents made three or more significant business changes at the same time, such as new products, new markets, more suppliers or more staff. Among those firms, 59% reported difficulty combining financial data, against 24% of firms making fewer changes. Reconciliation difficulty showed the same gap, at 59% against 27%.
The most telling number is how those companies hold the work together. Fifty-two percent of the high-change firms said they rely on manual or spreadsheet processes, compared with 6% of the steadier group. When PYMNTS asked respondents to rate what helped and what hurt, manual and spreadsheet processes were the only approach with a negative score, at minus 36 overall and minus 92 among the high-change firms. Standardizing processes scored plus 93, finance automation plus 91, data integration plus 88 and upgrading ERP systems plus 73.
That sequence matters for anyone trying to fix the forecast directly. A cash forecast is only as reliable as the reconciled balances, receivables data and system feeds it is built on. When a business adds a supplier base or a product line faster than finance can connect the data, the forecast inherits every gap, and the spreadsheet becomes the place where those gaps are patched by hand.
Treasury teams see the same pressure and are looking to AI to relieve it, but cautiously. A EuroFinance study supported by TIS, released in July, found that 38% of treasurers consider cash flow forecasting the most suitable use case for AI, and 46% are actively evaluating AI solutions, more than any other new technology. Only 4% said AI processes are already in production, while 47% know the forecasting use case but have no immediate plans to adopt it. According to TIS, more than half of treasurers identify data quality as the biggest drag on forecast accuracy. Robert McAnally, head of treasury and corporate finance at Siemens Energy, warned that teams "cannot ignore the data and assumptions behind AI-driven models."
The customer side of the forecast is getting harder to predict as well. In the October edition of PYMNTS Intelligence and J.P. Morgan's Time to Cash study, only 38% of CFOs said customer payment-term behavior improved, down from 73% a year earlier, even as 70% shortened their cash conversion cycle again. When collection timing becomes less predictable, the receivables assumptions in the forecast need to be measured by customer, not set once and assumed.
Playbook
Reconciliation difficulty more than doubles at firms making three or more changes at once. This playbook covers the close processes, reporting infrastructure and controls that keep finance accurate as the business adds lines and entities.
Download
Guide
Only 12% of mid-market finance leaders call their systems fully ready, and treasurers name data quality as AI's biggest obstacle in forecasting. This guide covers how to structure an ERP migration so connected data and AI are designed in from day one.
Download
Guide
A cash forecast is only as good as its collection assumptions, and only 38% of CFOs say customers are paying better. These AR metrics and benchmarks give the forecast measured days-to-pay instead of payment terms taken on faith.
Download
Mid-market CFOs shortened their cash cycles, yet far fewer say customer payment-term behavior improved.
CFOs say cash flow takes up more of their role, yet fewer are modeling scenarios as working capital strain spreads.
The Commercial Payments Bill would cap payment terms for large firms paying smaller suppliers and make late-payment interest mandatory.
