How much cash should your business actually hold? (A framework, not a rule of thumb.)
How much cash should a small business keep in the bank? The standard answer is three to six months of expenses, but that is a starting point, not an answer. Here is the honest framework for the right cash buffer for your specific business, accounting for seasonality, AR delays, and lumpy operating expenses.
A home-services owner we work with asked us a version of this question a couple of weeks ago: what is the right amount of cash we should hold in the bank to cover our seasonal swings and all the lumpy operating expenses we know are coming?
The standard answer is bad. Every finance textbook says three to six months of operating expenses. That is a starting point, not an answer. A software company with 90% margins and monthly recurring revenue has a completely different cash-buffer problem than a home-services firm swinging between winter peaks and summer troughs. The rule flattens all of these into the same number, and the number is wrong in most of them.
Here is the honest version. Five buckets that determine your right cash buffer, four questions that find your number, and a couple of specific signs you are holding too little or too much.
The five buckets a cash buffer covers
Your right total is the sum, not one of them. Each bucket depends on your business.
-
01
Normal monthly operating cash
The base layer. Enough to cover one full month of payroll, rent, software, insurance, and vendors even if zero revenue comes in. This is the only part the 3-to-6-month rule handles reasonably.
-
02
Seasonal-dip coverage
The bucket most operators forget. If revenue is seasonal, you need enough cash to cover the gap between your slowest months and your steady-state. On top of the base layer, not instead of. A home-services company doing $200K a month in winter and $60K in July needs to cover the $140K monthly gap for however many months the slow season lasts.
-
03
AR-delay worst case
Not average. Worst. Assume your two biggest customers each stretch payment by an extra 45 days at the same time. For most services businesses this is 60 to 90 days of billing tied up at any moment.
-
04
Known lumpy expenses
Not surprises, but frequently unplanned. Quarterly estimated taxes. Annual insurance renewals. Bonus payouts. Equipment purchases. Put every one on a rolling 12-month view and cover the biggest single hit that could land in any month.
-
05
Unknown-unknowns cushion
For the thing you did not see coming. Key employee leaves, vendor doubles price, truck gets totaled. Most operators land at 30 to 60 days of normal OpEx here.
The four questions that find YOUR number
How seasonal is your revenue? Pull the last 24 months. Find your three lowest consecutive months. The gap between that trough and your steady-state, multiplied by trough duration, is your seasonal-dip requirement.
What is your worst realistic AR delay? Not average. Assume your two biggest customers stretch payment by an extra 45 days simultaneously. If that puts you within striking distance of missing payroll, your AR bucket needs to be much bigger.
What is the biggest single expense hit that could land in any month? Insurance renewal, quarterly tax, bonus payout, equipment purchase. Your buffer covers it independently, because these events sometimes stack on top of a bad revenue month.
How fast could you adjust if things went badly? Variable-cost service firms flex down quickly. Capital-heavy businesses with fixed rent and salaried staff cannot. Slower adjustment means larger cushion.
Rough shapes by business type
Services (agencies, consulting). Two to four months of OpEx plus AR coverage for your biggest clients on 60-day worst-case terms.
Home services and trades. Three to six months of OpEx, plus the seasonal trough on top. JPMC data pegs the median construction SMB at 20 buffer days, which is not enough. If you are running at that level, one bad month puts you in serious trouble.
Product / retail / inventory-heavy. Four to six months, because inventory ties up cash and reorders are lumpy. JPMC median retail is 19 days, restaurants 16, both dangerously thin.
SaaS. Six to twelve months of burn pre-profit, three to six months if cash-flow positive.
The single most useful tool
A rolling 13-week cash forecast beats every rule of thumb.
The number every operator should see weekly: projected cash balance for each of the next 13 weeks, with every known outflow (payroll, rent, tax, vendor payments) and every projected inflow (AR receipts, expected new revenue) plotted forward.
The 13-week horizon is long enough to see the next quarterly tax hit and the next seasonal dip, short enough that projections are credible. Enterprise finance teams run this weekly. Most small businesses do not run it at all, which is why they get surprised.
Signs you are off
Too little. You have ever wondered whether payroll would clear. You are stretching vendors past their terms because the cash is not there. You are checking the account balance every Monday and doing mental math.
Too much. Cash sitting in a checking account earning zero when a treasury sweep would earn short-term Treasury rates. Not making a hire the team clearly needs. Not investing in equipment or marketing with a proven payback. Excess cash is opportunity cost, not free money.
A business that is running its finances well should not require the owner to worry about payroll. If yours does, the number in your bank account is not big enough for your business.
Sources cited:
- JPMorgan Chase Institute, "Cash is King: Flows, Balances, and Buffer Days" (analysis of ~597,000 U.S. small-business bank accounts, 2016)
- Bluevine Cash Flow Management Survey, September 2025 (n=774 U.S. small-business owners)
- Federal Reserve Small Business Credit Survey, 2025 Report on Employer Firms
- Porter customer conversation, July 2026, home-services owner (anonymized)
Common questions
01. How much cash should a small business keep in the bank?
There is no single number. The right total is the sum of five buckets: one month of normal operating cash, seasonal-dip coverage, worst-case AR delay coverage, the biggest single lumpy expense that could land in any month, and a 30-to-60-day unknown-unknowns cushion. For most services businesses this lands at two to four months of OpEx; for home services and trades, three to six months plus seasonal trough coverage; for SaaS pre-profit, six to twelve months of burn.
