How Much Startup Runway Do You Need if a Raise Fails?

Startup runway is cash divided by monthly net burn, and the common target of 12 to 18 months is sized for a raise that succeeds. Size it for one that fails: the months spent trying, then the cash a cut consumes before it saves anything. That sets a decision month well before the money runs out.

Carta’s data put the median gap between funding rounds, across all stages, at 696 days in Q2 2025, about 22.9 months (our arithmetic; Carta, 2025). In 2022 Sequoia advised its founders to start raising 12 months before the cash runs out, and to hold runway equal to the time to the next milestone plus 12 months (Gupta & Huang, 2022). That rule assumes the raise works, and aviation sizes its runways on the opposite assumption.

Exhibit 1. How much runway the guidance and the data imply

Source	Guidance or finding	Runway it implies	What it assumes
Common online guides (Excedr, Upsilon, Gynger)	12 to 18 months	12 to 18 months	No source given
Paul Graham (2014)	Raise at once with 8 months or less, before the “fatal pinch”	More than 8 months at all times	Investor interest falls as the cash runs down
Sequoia, Extending Your Runway (2022)	Raise 12 months before cash-out; hold the time to the goal plus 12 months	Goal + 12 months	The raise succeeds
Sequoia, Forecasting & Scenario Planning (2022)	At least 36 months, or fully funded	36 months	Funding stays scarce for about three years, as after 2000
Carta (Q4 2024; Q2 2025)	Median 696 days between rounds; 75th percentile above three years from seed to Series A	22.9 months for the median company; over 36 for one seed company in four	The last wait predicts the next
This piece	Months a failed raise lasts + stop distance of a cut + a 12-month reserve	About 16 months when the raise starts; about 34 after a round (Exhibit 3)	The raise can fail

Source: the Excedr, Upsilon and Gynger guides (3 October 2026); Graham (2014); Sequoia (2022); Carta (2025); for the last row, CX Cash’s method in Exhibit 3.

Startup runway divides net cash by net burn, and neither number holds still

Bessemer Venture Partners gives the standard calculation: a company can divide “its monthly net burn rate into a cash balance to understand its number of months of runway” (Bessemer). Net burn is cash paid out less cash collected; gross burn, the payments alone, gives the runway if revenue stopped. A hypothetical Series A software company with $15.0M of cash, paying out $1.0M a month and collecting $250k, has a startup runway of 20 months on net burn and 15 on gross burn.

Sequoia counts runway on net cash, so $10M in the bank with $5M of venture debt drawn is $5M (Gupta & Huang, 2022). Its partners added that “Runway is not static”: lumpy outlays can leave 3 months of it “when you thought it was 15”. The simulator below finds the month the balance reaches zero at a burn that can grow; the decision month falls well before it.

Rounds now arrive about two years apart, longer than 18 months of runway

The median company that raised a Series A on Carta in Q4 2024 had waited 774 days since its previous primary round, about 25.4 months (our arithmetic), the longest interval Carta had recorded this decade. From seed to Series A, the 75th percentile exceeded three years (Carta, 2025).

Bridges cover part of the gap: they made up about 40% of seed-stage rounds in 2024, against 36% in 2023 (Carta, Q4 2024).

The raise itself takes months. In DocSend’s 2023 seed data, half of the successful raises took 13 to 24 weeks (DocSend, 2023). Its pre-seed data say more about failure: unsuccessful companies kept raising for five months on average (DocSend, 2023). The samples are small (170 decks at seed, over 200 at pre-seed) and from the earliest rounds, so they give only an order of magnitude for a later raise.

Paul Graham’s essay “The Fatal Pinch” describes where a late start ends: a startup with, say, 6 months of runway and mediocre growth that expects to raise its way out. Graham put its chances dryly: “I will now, by an amazing feat of clairvoyance, do this for you: the probability is zero” (Graham, 2014). His rule was to raise at once with 8 months of runway or less. A company that follows the 12-month rule and spends five months on a raise that fails reaches that point with 7 months left, and with a cut still to pay for.

Aviation sizes a runway for the takeoff that is abandoned

Federal rules define V1 as “the maximum speed in the takeoff at which the pilot must take the first action” to stop the airplane within the accelerate-stop distance. Above V1, after an engine failure, the takeoff continues (14 CFR 1.2). The accelerate-stop distance adds the run to the engine failure, the run while the pilot reacts and the stop, plus a margin of “a distance equivalent to 2 seconds” at V1 (14 CFR 25.109). An airline may not take off unless that distance fits within the runway plus any stopway (14 CFR 121.189). The runway is sized for the takeoff that fails at the worst moment.

For a startup, the takeoff roll is the raise. V1 is the decision month: the last month in which the company can still cut its burn and stop on the cash it has left, with a reserve. The 2 seconds are the weeks a board takes to accept that a raise has failed. Call it the stop distance: the cash a company spends between deciding to cut and the first month the lower burn holds. Sequoia’s rule of raising with 12 months left measures the takeoff roll and leaves out the stop.

Three differences matter. A startup’s V1 is an estimate that drifts with each month of revenue, while an aircraft’s is computed from physics before every flight. An aborted raise leaves a smaller company, and a startup’s runway lengthens as it rolls if revenue grows.

But a cut spends cash before it saves any, so the decision month comes early

Sequoia’s March 2020 memo to its founders put the asymmetry in one line: “In downturns, revenue and cash levels always fall faster than expenses” (Sequoia, 2020). Expenses fall slowly for a plain reason: cutting them takes time and cash, so any plan to extend startup runway must start before the raise fails.

Severance is the largest one-off cost in the stop distance. When Stripe cut about 14% of its staff in November 2022, it paid 14 weeks of severance to everyone leaving, more for longer tenure (Stripe, 2022). Airbnb, cutting about a quarter of its staff in May 2020, paid US employees 14 weeks of base pay plus a week for each year of service (Airbnb, 2020). Fourteen weeks is about 3.2 months of the departing payroll (our arithmetic). Both were large companies; a young startup may pay less. People are also most of what a startup spends: in Kruze Consulting’s client data, payroll-related costs, contractors included, make up 76% of operating expenses (Kruze, 2024).

The law can add notice. The federal WARN Act requires 60 calendar days’ notice of a mass layoff from employers with 100 or more employees (20 CFR 639), and California’s version applies from 75 employees and to any layoff of 50 or more within 30 days (California EDD). The federal “faltering company” exception, for an employer actively seeking capital, covers plant closings and excludes mass layoffs. A 30-person seed company sits below every threshold; a 150-person Series B company in California may not.

Rank the levers that extend runway by when the cash arrives

PayPal’s response in 2000 shows why the order matters. Roelof Botha, then its chief financial officer, told Sequoia’s founders in 2022 that PayPal closed $100M at a $500M valuation in March 2000, a month before the Nasdaq fell; after it, “We had 7 months of runway left”. Michael Moritz’s advice, in Botha’s telling, was to “recognize the moment, expand runway, become a business”. PayPal ran cost exercises and introduced fees in June 2000 (Botha, Adapting to Endure, 2022). Fees were the lever with the shortest delay, since they applied to the next transaction.

Exhibit 2. Levers that extend runway, ordered by how soon the cash moves

Lever	When the cash first moves	What delays it	Evidence
Prices and fees on new business	Next transaction or invoice	Customer response	PayPal, June 2000
Contractors and consultants	End of each contract’s notice	Contract terms	12.5% of startup operating costs (Kruze, 2024)
Collections and annual prepayment	Next billing cycle	Customer terms; moves cash once, does not lower burn	Sequoia (2022)
Layoff	After notice and severance	14 weeks of severance; 60 days’ notice where WARN applies	Stripe (2022); Airbnb (2020); 20 CFR 639
R&D payroll tax credit	First quarter after the income tax return is filed	Filing date; offsets employer payroll taxes only	IRS; 26 U.S.C. § 41(h)
Undrawn debt line	When drawn, if ever	Lender approval and covenants; best left undrawn	Sequoia (2022)
New equity	13 to 24 weeks for half of successful 2023 seed raises	Investor appetite	DocSend (2023)

Source: as listed in the last column.

For a founder trying to lengthen startup runway, timing also decides what to prepare first. Doug Leone’s advice in the 2022 session was to do the cut exercise now so that the company is “ready to do it in the next 30 days if needed”. Preparation is the cheapest lever: it shortens the stop distance without cutting anything. Alfred Lin’s notes from the same session add the cost of cutting piecemeal: “Tellme - forced to cut (3x, better if 1x)”. Each round of cuts restarts the clock on notice and severance.

Collections move cash forward once, and the research credit moves it late. A young company with gross receipts under $5M can apply up to $500,000 a year of its research credit against payroll taxes (IRS; 26 U.S.C. § 41(h)). The credit starts in the quarter after the income tax return is filed and offsets only the employer’s 6.2% social security and 1.45% Medicare taxes (IRS). On $2.0M of quarterly wages, $500,000 takes about 3.3 quarters to absorb: cash for the second year of a runway plan.

The debt line belongs in a different column. Sequoia, which counts runway on net cash, recommends securing a line “and just not to consider it part of your runway” (Gupta & Huang, 2022). Aviation has a word for that kind of reserve: the stopway, an area beyond the runway “designated by the airport authorities for use in decelerating the airplane during an aborted takeoff” (14 CFR 1.1). The rules let a stopway count toward the distance needed to stop, never toward the takeoff run. An undrawn line deserves the same treatment: cash for the abort, not months for the raise.

Even the abort is not assured. Couchbase’s 2021 loan agreement with Silicon Valley Bank conditioned every advance on the bank’s satisfaction that no material adverse change had occurred (Couchbase S-1, Exhibit 10.13). Nerdy’s 2025 facility with Hercules leaves its second tranche to the lenders’ approval (Nerdy 8-K). A line can close in the month it is needed.

Therefore the raise should start with about 16 months of cash left, not 12

Cash needed at the decision month = stop distance + 12 × reduced monthly burn; latest raise start = decision month − months a failed raise lasts

Return to the hypothetical Series A company. Of the $1.0M it spends each month, 76% goes on people (Kruze’s average). Its prepared plan cuts people costs by 30% and other spending by 25%, which lowers net burn to $462k. Departing staff receive 14 weeks of pay, and the other savings take effect two months after the cut is announced. After the cut, the company wants 12 months of cash at the new burn, the line below which Sequoia called the runway problem “existential”. A failed raise is taken to last five months, DocSend’s pre-seed average; a Series A raise may fail faster or slower.

Exhibit 3. The decision month for the hypothetical company, $ thousands

Case	Stop distance	Cash needed to stop	Decision month	Months of cash left then	Latest start of a five-month raise
Prepared cut, no WARN notice	2,531	8,075	9	11.0	Month 4, with 16.0 months left
Prepared cut, 60 days’ WARN notice	2,987	8,531	8	12.0	Month 3, with 17.0 months left
Unprepared: two more months to plan	4,031	9,575	7	13.0	Month 2, with 18.0 months left

Hypothetical company as described above. Severance: Stripe (2022), Airbnb (2020). Notice: 20 CFR 639. Failed raise: DocSend (2023). Reserve: Sequoia (2022). Stop distance includes severance. Every figure is computed from these inputs.

In the prepared case the stop distance is $2.5M, or 3.4 months of today’s burn, and the company must decide by month 9, with 11 months of cash left. A raise that may take five months to fail must therefore start by month 4, with 16 months left. Under the 12-month rule the raise starts at month 8 and its failure becomes clear at month 13. The cut then leaves $2.7M, or 5.9 months at the reduced burn, inside the bucket Sequoia called existential. Preparation is worth two months of decision time, since the unprepared company must decide at month 7.

The same logic sizes the round: suppose Carta’s median interval of 22.9 months holds and the raise takes 5 of those months. The company then reaches its milestone 17.9 months after closing and must still hold 16 months of cash, so it needs about 34 months of startup runway when the previous round closes. Sequoia asked for 36 in 2022, and its forecasting session gave the reason: “during the Internet bubble, it took about three years before it felt like we were back to a reasonably normal funding environment” (Gupta & Grady, 2022).

But the cycle decides whether to raise before the milestone or after it

The decision month fixes the latest start; the cycle decides how much sooner to go, because it sets what investors pay for. In booms they pay for promise: Paul Gompers and Josh Lerner found that fund inflows raise the prices of new investments, changes that “do not appear related to the ultimate success of these firms” (Gompers & Lerner, 1998). In recessions they pay for proof. In four decades of US data, early-stage deal counts fall about 30% “while late-stage activity is relatively unaffected”, and follow-ons take a larger share of deals (Howell et al., 2023).

Exhibit 4. Raise now or after a milestone six months away, hypothetical Series A, $ millions

Market now, then in six months	Pre-money now	Ownership sold now	Pre-money after proof / a miss	Ownership sold after proof / a miss	Waiting pays if the chance of proof exceeds
Boom, still a boom	40	16.7%	56 / 32	15.7% / 24.5%	89%
Bust, still a bust	16	33.3%	36 / 14	22.4% / 42.6%	46%
Boom, turned to bust	40	16.7%	36 / 14	22.4% / 42.6%	Never

Source: CX Cash calculation on assumed valuations; the company raises $8.0M now, or $10.4M after six months at $0.4M of net burn. Ownership sold = raise ÷ (pre-money + raise).

In a boom that lasts, waiting pays only above 89% odds, and a certain milestone saves one point of dilution. In a bust the milestone is worth about 11 points, and waiting pays at better than even odds. After the 2022 reset, the median wait from seed to Series A went from 420 days in Q4 2021 to 774 three years later (Carta, 2025). A round raised on promise therefore has to reach the milestone in a cold market, as the 34 months above would. In a bust the raise should start once the milestone can be checked in diligence; deals took 83 days on average to close in a survey of 885 venture capitalists (Gompers et al., 2016).

Carta says the 2022 hangover is over

The best case against a long runway is the market’s recovery. Carta’s data for Q4 2025 show startups raising $36.1B, the most in any quarter since Q2 2022, with down rounds below 14% of new rounds. In Carta’s words, “The hangover from the market reset that occurred in 2022 is over” (Carta, 2026). Sequoia’s Alfred Lin warned in 2022 that “You won’t be able to cut your way to a business”. A company that cut to 36 months of runway in 2022 may have given up growth that the 2025 market would have paid for.

Lin’s warning holds: a cut that removes the growth engine moves the milestone further away. Markets do reopen, and by Carta’s numbers this one took about three years, close to what Sequoia’s 36 months assumed.

The recovery argument says less about who the reopening reached. The same report counts 4,859 new rounds in 2025, the lowest annual total in at least six years and 41% below 2021. The median wait between rounds was still rising in 2025. And the decision month does not require cutting early. It requires a prepared cut and a date.

Put the decision month on the board agenda

If the next raise is less than two years away

• Compute the decision month beside the zero-cash month: the last month in which a prepared cut still leaves 12 months of cash at the reduced burn.
• Start raising at least five months before it, sooner in a boom, and size that round for a bust.
• Price the cut now: 14 weeks of severance, 60 days of notice where WARN applies, and the renewal date of every large contract.
• Report the decision month in each monthly investor update; the template below is a starting point.

When the board sets the trigger

• Ask for the decision month beside the runway, with its assumptions, and whether the cut plan could run within 30 days, as Leone advised.
• Read the cycle from the pipeline: term sheets offered before diligence point to a boom, and interest only from existing investors to a bust.
• Count startup runway on net cash, and leave an undrawn debt line on the stopway.
• Agree the trigger in advance: a signed term sheet by the decision month, or the prepared plan runs.

Runway sized for the raise that fails

How much startup runway a company needs depends on what happens if its next raise fails. Aviation’s method answers it: price the stop distance, keep a reserve, start the raise before the decision month and leave the debt line on the stopway. For the hypothetical company in Exhibit 3, the raise starts with about 16 months of cash left, 4 months earlier than the 12-month rule.

The runway you need is the one you can still stop on.

Notes and sources

• Airbnb (2020). A Message from Co-Founder and CEO Brian Chesky, 5 May.
• Bessemer Venture Partners. The five accounting metrics for cloud companies.
• Botha, R. (2022). Why We’re Here, in Adapting to Endure. Sequoia Capital presentation to founders, 16 May; published 14 June.
• California Employment Development Department. Worker Adjustment and Retraining Notification (WARN): federal and California WARN compared.
• Carta (2025). State of Private Markets: Q4 and 2024 in review; Dowd, K., The typical time between VC rounds is shrinking in SaaS and rising in fintech, 19 March; Neville, A. and Dowd, K., State of Private Markets: Q2 2025, 18 August.
• Carta (2026). Neville, A., State of Private Markets: Q4 2025.
• DocSend (2023). From FOMO to JOMO (pre-seed report), press release, 24 August; The seed round in 2022-23, press release, 7 December.
• Gompers, P. and Lerner, J. (1998). Money chasing deals? The impact of fund inflows on private equity valuations. SSRN working paper; published in the Journal of Financial Economics 55(2), 281–325 (2000).
• Gompers, P., Gornall, W., Kaplan, S. N. and Strebulaev, I. A. (2016). How do venture capitalists make decisions? NBER Working Paper 22587; published in the Journal of Financial Economics (2020).
• Graham, P. (2014). The Fatal Pinch. paulgraham.com, December.
• Gupta, R. and Grady, P. (2022). Forecasting & Scenario Planning. Sequoia Capital presentation to founders.
• Gupta, R. and Huang, S. (2022). Extending Your Runway. Sequoia Capital presentation to founders, 26 May.
• Howell, S. T., Lerner, J., Nanda, R. and Townsend, R. R. (2023). How resilient is venture-backed innovation? Evidence from four decades of U.S. patenting. NBER Working Paper 27150 (first circulated 2020).
• Internal Revenue Service. Qualified small business payroll tax credit for increasing research activities; Topic no. 751, Social Security and Medicare withholding rates.
• Kruze Consulting (2024). Startup Payroll Costs and Tax Budgeting for New Hires, 5 November.
• Leone, D. and Lin, A. (2022). What History Tells Us and How to Emerge Stronger, in Adapting to Endure. Sequoia Capital.
• Sequoia Capital (2020). Coronavirus: The Black Swan of 2020, 5 March.
• Stripe (2022). CEO Patrick Collison’s email to Stripe employees, 3 November.
• US Code of Federal Regulations: 14 CFR 1.1 and 1.2 (definitions), 14 CFR 25.109 (accelerate-stop distance), 14 CFR 121.189 (takeoff limitations), 20 CFR Part 639 (WARN). eCFR.
• US Code: 26 U.S.C. § 41(h), credit for increasing research activities (qualified small businesses).

The Sequoia quotations come from the speaker notes in its published presentations, and DocSend’s figures from press releases on small samples, used here only to size a hypothetical failed raise. Where a figure has no source beside it, CX Cash worked it out from the cited numbers. The companies in Exhibits 3 and 4 are hypothetical; the cut sizes, the two-month lag on non-people savings and the Exhibit 4 valuations are assumptions. Gompers and Lerner’s finding comes from the paper’s abstract, and Carta’s figures cover companies on its platform. “Stop distance” is CX Cash’s term.