Why Your Cash Flow Model Is Lying to You.

Your cash flow model was probably built in Excel, inherited from your accountant, and last seriously challenged when you were raising funding. It shows you a runway number. It shows you a break-even date. It gives you confidence. It is almost certainly wrong.

Not wrong in the sense of bad maths. Wrong in the sense that it is built on assumptions that were always optimistic, were never tested against reality, and systematically understate the speed at which a business can run out of options.

We've reviewed cash flow models for over 40 SMBs in the past two years. The same three false assumptions appear in almost all of them. Here they are, with the corrections that actually matter.

82% of SMBs that fail cite cash flow problems as the primary cause — not market conditions. 4–6 months is the typical gap between when a cash crisis becomes visible and when it became inevitable. And there is 1 number that actually predicts your survival — and it's not runway.

Your model shows revenue appearing in the month it's billed. In reality, cash arrives when your customer pays — which is typically 30, 60, or 90 days after invoice. Enterprise customers with 90-day payment terms can create a structural funding gap that your model treats as a cash flow event but your bank account treats as a future promise.

The compounding problem: when you're growing, this gap widens. If you bill £100,000 in January on 60-day terms, you receive £100,000 in March. If you bill £150,000 in February, you receive that in April. Meanwhile your costs — salaries, software, office — arrive monthly, on time, without exception. A growing business with long payment terms can show strong revenue growth and deteriorating cash simultaneously. Most cash flow models smooth this over entirely.

The fix: build a separate collections timeline into your model. Map each revenue source to its actual average payment date, not invoice date. Then stress-test with a 20% payment delay scenario.

Your model has a cost structure that looks stable month to month. Perhaps a 5–10% buffer above your known fixed costs. What it doesn't model is the reality that SMBs routinely encounter large, semi-predictable costs that are routinely absent from base case models: insurance renewals, software contract true-ups, one-time legal fees, equipment replacement, redundancy costs, or a customer who demands an emergency SLA response that you under-resourced for.

None of these are truly unpredictable. All of them happen. The average SMB experiences at least two significant unplanned cost events per year that are 3–5x larger than their monthly buffer. If your model has an 8% cost buffer and you face a £40,000 legal dispute, an emergency infrastructure upgrade, and a key hire to replace an unexpected departure in the same quarter, your model's buffer is exhausted within weeks.

The fix: build a separate "shock" reserve distinct from your operating buffer — funded to cover at least one large event — and model what happens to your runway if two medium events occur in the same quarter.

"Cash flow models are built to secure funding and reassure boards. They are rarely built to actually predict when you run out of options."

The most dangerous assumption in most SMB cash flow models is that growth arrives smoothly and costs scale proportionally with it. In practice, growth is lumpy — large new customers or contracts arrive in clusters — and costs often scale ahead of revenue, not with it. You hire the team before the customer renews. You invest in infrastructure before the pipeline converts. You sign the office lease before the headcount arrives.

This creates a predictable pattern: cash dips sharply every time the business invests ahead of growth, recovers when the growth materialises, then dips again on the next investment cycle. Businesses that model smooth linear growth consistently find themselves surprised by the depth of the dips. They know the recovery is coming, but the dip arrives faster and deeper than modelled, and if cash reserves are insufficient, the recovery is irrelevant.

The fix: model your growth in cohorts — when does each new customer or contract actually pay? Build a step-function cost model that shows costs jumping at hiring thresholds, then flat until the next threshold. The shape of your cash flow curve changes entirely — and becomes far more useful.

Runway — months of cash remaining at current burn — is the metric everyone watches. It is also the metric that gives false comfort. A business with 12 months of runway and accelerating burn has less time than its runway suggests. A business with 6 months of runway and decelerating burn has more.

The number that actually predicts survival is Cash Conversion Efficiency (CCE): the ratio of cash collected to cash consumed, measured monthly and trended over time.

CCE = Cash Collected (month) ÷ Cash Consumed (month). Target: CCE trending toward 1.0 or above. A CCE below 1.0 means you are consuming more cash than you are collecting — you are burning. A CCE above 1.0 means you are generating cash. The trend of CCE over three to six months tells you more about your survival probability than any single runway number. A business with 18 months of runway and a CCE declining from 0.9 to 0.7 over six months is in more danger than one with 8 months of runway and a CCE rising from 0.7 to 0.95.

Track CCE monthly. Plot it as a trend. If it is declining, understand exactly why before you spend another pound on growth. If it is improving, understand what is driving that improvement and protect it.

The reason cash flow crises are so often described as sudden is that most businesses detect them late. The model shows runway. The bank account looks acceptable. The revenue pipeline looks healthy. And then, usually in a single bad quarter where payment delays coincide with a cost shock and a growth investment, the situation becomes urgent faster than anyone modelled.

The window between "this is a problem we need to address" and "we have run out of options" is typically four to six months for an SMB. Fundraising takes three months minimum. Asset sales take longer. Cost restructuring has a minimum lead time before cash impact. That window is smaller than most founders think, and the model that's supposed to give early warning is the same model that failed to flag it.