Direct vs indirect cash flow forecasting: which method fits a mid-sized company?

The short verdict

Choose the direct method when the question is whether the company has enough cash over the next few weeks, and the indirect method when the question is what the business will generate over the next year. A mid-sized company with a lender, a seasonal working-capital swing or a tight quarter usually needs both: a weekly direct forecast covering about 13 weeks, rolling into a monthly indirect forecast tied to the budget and the three-statement model.

The direct method forecasts actual cash movements line by line: customer receipts, payroll, vendor payment runs, rent, taxes and debt service. The indirect method starts from forecast net income and works back to cash, adding back non-cash charges such as depreciation, then adjusting for changes in receivables, inventory, payables and other balances, plus capex and financing flows.

Direct vs indirect: side-by-side comparison

When the direct method wins

The direct method wins whenever timing inside the month matters more than the month-end total.

• Cash is tight, or a covenant tests liquidity rather than earnings.
• A lender, sponsor or turnaround adviser asks for a 13-week forecast with weekly variance reporting.
• Receipts are lumpy: milestone billing, a few large customers, slow public-sector payers or seasonal collections.
• Payroll, tax deposits and debt service cluster in particular weeks.
• You need to decide which vendors get paid this week.

The method is only as good as the customer payment history under it. A receipts forecast that applies each customer's actual days-to-pay, learned from several years of invoices, beats one that applies a blended DSO to the whole ledger. That means invoice-level AR with issue dates, due dates, payment dates and credits, and AP with terms and actual payment dates. If receipts beyond open invoices come from pipeline, the guide to building a sales forecast from CRM deal history covers converting won deals into billing and collection dates.

When the indirect method wins

The indirect method wins when the forecast has to agree with the P&L and balance sheet over a longer horizon.

• The forecast must run 12 to 36 months for a budget, refinancing or acquisition model.
• The board wants cash reconciled to EBITDA and the balance sheet.
• Working capital follows stable ratios rather than a handful of large invoices.
• You are testing scenarios such as price changes, hiring plans, capex or a new site.
• A valuation, earnout or covenant model needs consistent three-statement outputs.

Its weakness is timing. A monthly indirect forecast can show a comfortable month-end while payroll and a loan payment land in week two and the largest customer pays in week four. That gap is why the methods are usually run together.

How to run both without doing the work twice

1. Build the direct forecast for weeks 1 to 13 from open AR, open AP, the payroll calendar, tax dates and the debt schedule.
2. Build the indirect forecast monthly from the P&L model, starting with the first month after week 13.
3. Reconcile the overlap: the direct forecast's month-end cash for the first three months should sit close to the indirect forecast for those months. Explain any difference by line.
4. Feed variances back. When actual receipts beat or miss forecast, update customer payment assumptions in both models.
5. Keep one-off items, such as asset sales, insurance recoveries, tax refunds or license receipts, as named lines in both forecasts rather than burying them in working capital.

For a client with several entities, build cash by entity before building the group view. The page on consolidating multiple QuickBooks company files covers the account mapping and intercompany eliminations that come first.

Your AR and AP history is also a records-depth signal

The first time you build a direct forecast for a client, you learn how deep their transaction history goes. That matters beyond treasury. A company that can produce years of invoice-level receivables and payables, tied to clean customer and vendor records, usually runs a disciplined back office, and the same discipline tends to show up in its CRM, support and operations systems.

Run this check while you build the receipts model:

• Can the client export invoice-level AR with issue, due and paid dates for at least three years?
• Are credits, disputes and write-offs recorded against the original invoices?
• Do customer IDs in the ledger match the CRM and the help desk?
• Are purchase orders and AP approvals kept in a system rather than inbox threads?
• Is any of that history at risk from a planned migration or subscription change?

Mostly yes answers suggest the company keeps the kind of connected operational records that AI labs and data buyers license, provided it also meets the size, history, rights and sponsor tests. Treat it as a prompt for a conversation, not a conclusion; the company fit checker gives a preliminary, non-binding read.

Where a possible license receipt belongs in the forecast

Keep it out of the base case until the agreement is signed. A data license brings a single one-time payment, and until the company has agreed price and terms and signed, there is nothing firm to forecast. After signing, the money typically arrives within about 60 days of invoicing once the buyer selects the data.

The 13-week forecast guide for one-time receipts covers the line-item mechanics, and the client's risk register is the right place to track the opportunity before it is signed.

Accounting timing can differ from cash timing. Under ASC 606, a license of intellectual property is assessed as either a right to use the IP as it exists when granted, recognized at a point in time, or a right to access it throughout the license period, recognized over time; Deloitte's ASC 606 licensing roadmap explains how that distinction is drawn. Because the indirect forecast starts from net income, any gap between recognized revenue and cash received appears as a receivable or deferred revenue movement, so ask the company's auditors how they would treat the specific agreement. This is general information, not legal, tax or financial advice.

How partner rewards work for a fractional CFO

Partners earn 25% of the eligible platform fees SourceX actually collects from the referred company's licensing deals, capped at $100,000 per referred company. Rewards become payable only after the buyer pays and SourceX receives its fee; a lead, meeting or signed agreement alone does not trigger payment.

For forecasting purposes, treat your own reward the way you would treat the client's receipt: not in any plan until it is paid. It comes out of SourceX's fee and never reduces what the client receives. Credit goes to the first valid referrer whose introduction leads to a verified company application within the attribution window, so date your introductions. If you hold a CPA license, read your state board's commission and referral fee rule before accepting anything; the partner program overview for fractional CFOs explains how attribution and payment work in practice.

Next step

The next time you build a 13-week forecast, run the records-depth check alongside it. If the client clears it and the owner is interested, register as a partner and make the introduction, or send the CEO to sourcex.si/apply with your referral link.