Cash flow & AR: see the next four weeks

Why four weeks

Beyond a month, uncertainty explodes for many SMBs; under a week, you are only firefighting. Four weeks is a horizon where issued invoices, payment promises, and upcoming payroll / tax remittances are still actionable. The cockpit then shows a projected liquidity band, not only yesterday’s bank balance.

The goal is not enterprise-grade treasury precision on day one. It is a shared truth good enough to decide: whom to chase, what to defer, where to brake.

AR: the most underestimated lever

Accounts receivable often hide invisible risk: invoices at +60 or +90 days, untracked disputes, strategic customers stretching terms. AR aging linked to accounts and payment promises feeds the forecast. AI can flag drift (rising DSO, risk concentrated in three accounts) and prepare the action list for finance leadership.

• AR aging by customer and age bucket
• Payment promises with credible dates
• Risk concentration (top accounts)
• Link to backlog and shipments to bill

Disbursements and predictable shocks

Payroll, tax remittances, critical supplier payments, debt service: these shocks are often known. Folding them into the forecast avoids “surprises” that are not surprises. Three-way match on the payables side (PO, receipt, invoice) also reduces leaks before cash out — a common Génosa use case.

From fragile spreadsheet to cockpit signal

Many teams still run cash in a brittle file. The healthy path: stabilize definitions, wire ERP / bank / AR into a mart, expose the four-week view in the leadership cockpit, then add thresholds and alerts. ChatGPT, Claude, Gemini, or Grok, connected to your data through Génosa, can summarize gaps for the committee — with data staying local or on a private network.

Recommended ritual

A short weekly cash review (15–20 minutes) with finance and the GM: variance vs last week, top AR risks, chase or defer decisions. Once a month, recalibrate assumptions. The free AI diagnostic helps measure the gap between today’s spreadsheet and a fundable cockpit.

Building the liquidity band

A useful forecast shows at least: opening balance, likely collections (AR × recovery factors by age bucket), known disbursements, and a safety buffer. Colour the band if the committee’s liquidity floor is threatened within 14 or 28 days. That signal, more than the chart itself, triggers action.

Calibrate recovery factors with real history, not sales optimism. Review monthly. Payment promises need a date and an owner; otherwise they stay out of the base case.

Integrating bank, ERP, and exception queues

The cash mart ideally consumes bank transactions, the general ledger, AR/AP aging, and payroll/tax schedules. AP exceptions (price or quantity gaps) delay payments or create duplicates — three-way match reduces noise before forecasting even begins.

The AI assistant can prep the review: top 10 at-risk AR, variance vs last week, three questions for the committee. It does not call the customer for you without an explicit mandate and human validation.

Common mistakes

Confusing bank balance with the ability to meet the next 30 days. Ignoring concentrations (one customer = 40% of AR). Updating the forecast only in a crisis. Or buying a complex tool before definitions are stable. The Génosa path remains: diagnostic, foundations, cockpit, signals.

Frequently asked questions