Why Your Startup Runway Calculation Is Probably Wrong

Your board deck says 14 months of runway. Your bank account says ten. The board deck is the one founders quote in fundraising conversations. The bank account is the one that ends payroll.

The two numbers come from the same company on the same day. They disagree because the formula on the slide is too simple to be right. Cash divided by trailing-three-month burn is the calculation every founder learns in their first month of running a startup. It fits on one line of a deck. And for most companies in their first 24 months, it is wrong by two to four months.

The trailing-burn formula assumes next month looks like last month. For an early-stage company, next month almost never looks like last month. Headcount is changing. Revenue is lumpy. A federal estimated tax payment is due in June. The Stripe payout you were counting on lands six business days later than the spreadsheet thinks. Each one of those is a small distortion. Stacked together, they move your zero-cash date by a full quarter.

This post is the runway calculation I walk founders through when their banker says 14 months and their controller says nine. The math is not hard. The discipline is catching every input that the simple formula ignores.

One thing upfront. Most founders do not need a fractional CFO to run this. A good controller and a 13-week cash forecast template will get you 90% of the way there. I will come back to that.

The two-line formula that fools everyone

The formula every founder learns:

Runway (months) = Cash on hand / Monthly net burn

It is right enough to be dangerous. Both inputs look obvious. Cash on hand is the number in your bank account. Monthly net burn is what the books say you burned last month, or the average of the last three.

What goes wrong with cash on hand. The bank balance is not the same as available cash. Available cash is operating bank balance, plus money market or treasury accounts, minus restricted cash (security deposits, escrow, pledged collateral), minus pending wires that have not cleared, minus the AP run you authorize tomorrow morning. The number in the bank account on Monday is rarely the number you actually have available to spend on Tuesday.

What goes wrong with net burn. Last month might have been a slow stretch for one-time vendor payments. Or it might have been the month your annual cybersecurity insurance renewed for $42,000 in a single hit. Either way, "last month" is a poor proxy for "next month."

Founders who anchor on the simple formula end up surprised every quarter when their actual cash position diverges from the model. The model is not wrong because the math is wrong. The inputs are.

Trailing burn versus forward burn

The single most common error: using trailing-three-month actual burn to project forward, when the company is growing.

If you hired three engineers in March, your March-through-May average burn does not reflect the burn rate those engineers create at full salary. It reflects partial months and a ramp. By July, when all three are at full cost, your actual burn is materially higher than the trailing average suggested.

The reverse fails too. If you raised prices in February and your trailing window is January through March, you are blending an old gross margin structure into a forecast that should reflect the new one.

The fix is to build forward burn from the bottom up. Take current headcount at full cost. Add committed-but-not-yet-active expenses (the sales hire who starts in eight weeks, the office lease that begins next month). Subtract contractor relationships that have wound down. The result is forward burn, which is what your runway calculation needs.

For a 40-person startup, the gap between trailing burn and forward burn is regularly $50,000 to $100,000 per month. That is one to two months of runway difference on a $1 million cash balance. A founder making a hiring decision off the trailing number is making it with the wrong information.

Net burn versus gross burn (and which one matters when)

Gross burn is total outflows. Salaries, rent, software, vendors, taxes. Every dollar going out.

Net burn is gross burn minus revenue collected. This is the number most founders use.

Net burn is fine for healthy companies with predictable, diversified revenue. It is dangerous for early-stage companies with lumpy or single-customer revenue.

A startup at $40,000 monthly gross burn and $25,000 monthly revenue has $15,000 net burn. With $300,000 of cash, the net-burn runway calculation says 20 months.

Now imagine the single largest customer (40% of revenue) churns next quarter. Revenue drops to $15,000. Net burn jumps to $25,000. Runway drops to 12 months. The 20-month number was real. It was also a coin balanced on one customer.

Model runway against both. Net-burn runway is the cheerful number for your board. Gross-burn runway is your worst-case floor: what if every dollar of revenue disappeared tomorrow. For the company above, gross-burn runway is 7.5 months. That is the number you fundraise against, because the investors will model it that way.

Run both. Quote net-burn to your board if they want optimism. Plan against gross-burn when you are deciding whether to hire.

The cash timing problem

The runway formula assumes cash flows evenly. It does not.

Some specific timing distortions that move runway by weeks:

Most startups pay employees on the 1st and 15th, and receive customer payments on a 30 to 60 day lag from invoicing. A month where you sent most invoices on the 25th and payroll runs on the 1st creates a temporary cash crunch even when the underlying business is fine.

Federal estimated tax payments are due quarterly (April, June, September, January). For a startup with net income (rare but it happens), these can be 20% of operating cash flow in those specific months. A naive runway calculation averages them across 12 months and misses the lumpiness.

Annual software contracts and insurance renewals create predictable one-time outflows. Cyber insurance, D&O insurance, the annual seat-license for your CRM. Five or ten of those stacked into one month look like a month where burn doubled. They are not changes in burn rate. They are timing artifacts the simple formula buries.

State franchise tax (Delaware C-corps) is due March 1. It is small for most startups ($400 to $200,000 depending on authorized shares and method elected) but easy to miss in modeling.

Customer collections on annual contracts. A SaaS company with a December renewal cycle collects most of its ARR in January and February. The cash flow profile of that company in March looks dramatically different from the same company in July. The annualized "monthly burn" number does not show this.

A real runway calculation accounts for these at weekly granularity for the next 12 weeks and monthly granularity from month four through month 18. Not for show. Because if you miss a payroll in week six, the company is done regardless of what month 14 looks like on the spreadsheet.

The burn multiple sanity check

David Sacks at Craft Ventures published the Burn Multiple framework in 2020. The version that matters:

Burn Multiple = Net Burn / Net New ARR (same period)

A burn multiple under 1.0 is efficient growth (you are creating more ARR than you are burning). 1.0 to 2.0 is acceptable. 2.0 to 3.0 is concerning. Over 3.0 is broken unit economics or a product that has not found its market.

Why this matters for runway. A startup with 18 months of trailing-burn runway and a burn multiple of 4 is in trouble even when the cash math looks fine. The cash will run out before the next raise is plausible, because investors will see the burn multiple, conclude the unit economics are broken, and pass.

A startup with 12 months of trailing-burn runway and a burn multiple of 0.8 is in better shape. The next raise is fundable.

Burn multiple is a quality metric stacked on top of runway. Run both. A pretty runway number with an ugly burn multiple is a fundraising problem disguised as a cash problem.

What "real" runway calculation looks like

The discipline that runs inside a tight finance function:

A 13-week cash forecast, updated weekly. Every line item that will hit the bank in the next 13 weeks, by week. Payroll, rent, vendors, customer collections, tax payments, one-time hits. Compared against the prior week's forecast. Variance explained. This catches timing problems before they become surprises.

A 12-month operating model with monthly granularity. Built bottom-up from current headcount and committed costs, layered with realistic revenue assumptions, run against three explicit scenarios (base, downside, upside).

Reconciliation of the two. The 13-week and the 12-month should agree on weeks one through 13 in the cash position they project. When they diverge, one of them is wrong. Usually the operating model.

For most startups under 30 people, this entire system fits in a single Google Sheet that takes two hours per month to maintain after the initial build. The real cost is the discipline of updating it weekly, not the cost of building it.

A startup that runs this discipline knows its runway to within plus or minus two weeks at any moment. A startup that does not is quoting a board-deck number that is off by months.

A note on AI in 2026

AI-driven cash forecasting tools have improved meaningfully in the last year, especially for companies with clean ERP data. Several can produce a defensible 13-week forecast from QuickBooks data automatically and reduce the manual maintenance work a lot.

What the tools cannot do: identify which customer contract has actual collection risk, decide whether to include a not-yet-signed term sheet in the upside case, or call your largest vendor and renegotiate net 60 to net 30 because cash is tight. Those are still founder decisions. The tools handle arithmetic. Judgment is yours.

What founders typically need (and what they do not)

If your books are clean, your headcount is stable, and your revenue is predictable, you can run this calculation yourself over a weekend. The tools and templates are not the bottleneck. The data quality is.

If your books are not clean, you do not need a runway tool. You need a controller. A real controller who closes the month in five business days, reconciles every account, and produces a P&L you can trust. Without that, every runway number downstream is built on sand. I wrote about this at length in why most startups hire a fractional CFO too early. The same principle applies here. Data integrity sits underneath every strategic finance question. Getting it right is a controller problem before it is a CFO problem.

I have turned away founders who wanted me to build them an integrated 13-week and 12-month model on QuickBooks data that was four months behind and missing two of their three bank feeds. The right answer was not me. The right answer was a $2,500-per-month bookkeeper and a $3,000-per-month part-time controller for 90 days to fix the foundation. Then come back if you still need help with the model.

If your revenue is recurring, your gross margins are stable, and you are inside the 12-to-24-month window before a Series A, your runway calculation is one of the three or four numbers that should be tight. The other two or three are gross margin, net retention, and burn multiple. The investor diligence call will start there.

Three things you can do this week

Take the Raise-Ready Scorecard if you are within 12 months of a raise. It tells you which of the runway and burn diagnostics you need to clean up before you start meeting with investors. Most founders learn they are six to nine months out from being raise-ready, which means you have time to fix the inputs before someone is doing diligence on them.

If you want a template you can install over a weekend, the CFO Toolkit includes a 13-week cash forecast and a 12-month operating model. Both are calibrated to early-stage startups, both are simple enough to maintain in a spreadsheet, and both account for the timing distortions this post covered.

If your runway feels off and you cannot explain why, book a 30-minute call. Bring your last bank statement and your last close. I will help you find whether your problem is the math, the data, or the inputs. Sometimes it is one of the three and you can fix it yourself in a week.

A runway number you cannot defend in a pitch is a meeting you have already lost.