How to Calculate Runway Under Revenue Volatility (the Honest Way)

Most runway calculators assume revenue is a constant number. Solo founders rarely have constant revenue. A consulting practice swings from $25k to $5k month-to-month. A subscription product loses one anchor account and drops 18%. The deterministic runway number hides the variance that actually drives going-concern risk. This article lays out the honest method, the three statistics that matter, and where the standard tools still earn their keep.

1. The naive runway formula and why it lies

The standard formula, covered in our runway planning article, is Runway = Cash / Net Monthly Burn, where net burn is gross burn minus collected revenue. It works when revenue is stable. It fails in a specific way when revenue varies: it computes the runway you have on average, not the runway you have with high confidence.

The lie is in the substitution. Replacing a volatile revenue stream with its mean quietly assumes that good months and bad months cancel out. They do not, because cash is path-dependent. A run of three bad months in months 4, 5, 6 forces a faster cash drawdown during a period when you have less cash than the starting balance. The same three bad months in months 10, 11, 12 are far less dangerous because cash has either accumulated breathing room or the company is already in action mode.

Concretely, with $600k cash, $100k gross burn, and revenue averaging $60k per month, the naive runway is $600k / $40k = 15 months. If revenue is drawn from a distribution with $60k mean and $18k standard deviation (30% CV), the probability that cash hits zero before month 15 is around 50%. Roughly half the time, you run out earlier than the calculator says you will.

Two failure modes follow. Founders trigger action at runway thresholds calculated against the mean, which means they trigger late in bad scenarios. Fundraising conversations anchored on the mean look fine on the deck and unconvincing once an investor pulls historical revenue and computes the variance themselves.

2. Three real revenue-volatility patterns

Volatility has different shapes depending on the business, and the shape changes which simulation assumptions hold. Three patterns cover most solo-founder cases.

Subscription churn drift. A SaaS or membership business with monthly subscription revenue. MRR looks smooth on a trailing chart, but month-over-month change is dominated by gross churn variability. OpenView's 2024 benchmarks report median SaaS gross retention of 89% with a 6 percentage-point standard deviation across the cohort^[3]. CV on monthly net new MRR for a small SaaS commonly sits in the 40 to 80% range, even when the headline ARR line looks tame.

Project-based feast/famine. Consulting, freelance, and agency revenue. Monthly revenue is the sum of a small number of project payments, so month-to-month variance is structural, not noise. A practice billing $30k average monthly revenue across 4 to 6 active engagements typically has a revenue standard deviation around $10 to 14k, putting CV at 33 to 47%. The pattern goes bimodal in the worst cases: months with a project kickoff or final invoice are double the average, months between projects a third of it.

Freemium conversion drift. Tools or content products with free-to-paid conversion as the primary growth driver. Top-of-funnel traffic varies with seasonality, search-algorithm changes, and content-publishing cadence. Conversion rates drift with audience composition. Monthly new-paid-customer counts have CVs in the 25 to 50% range even for businesses with stable funnels.

Knowing the pattern tells you what input the monte-carlo needs. Subscription businesses simulate net-new MRR. Project businesses simulate monthly revenue directly with a long-tail distribution. Freemium businesses simulate top-of-funnel traffic and conversion rate separately, then multiply.

3. The honest formula: monte-carlo runway with revenue CV

The honest method is a monte-carlo simulation. The structure is simple enough to fit in a spreadsheet, though most founders build it in Python or use a calculator that handles it. The recipe:

1. Compute revenue CV from history. Take trailing 12 months of cash-collected revenue. Compute mean (μ) and standard deviation (σ). CV = σ / μ. With less than 12 months, fall back to 30% as a conservative default; the Federal Reserve's 2024 SBCS data shows roughly 35% of small employer firms report substantial revenue variability quarter-to-quarter^[2].
2. Pick a distribution. Lognormal is the default for revenue (cannot go negative, right skew). Subscription net-new MRR fits a normal around the trend. Project-based revenue with a known billing cadence may need a custom mixture distribution.
3. Simulate N months of revenue. Draw N independent samples (or correlated, with evidence of momentum) for each forward month, for K trials. K = 10,000 is plenty.
4. Compute cash trajectory per trial. Walk the months: starting cash + simulated revenue − gross burn = end-of-month cash. Record the first month where cash crosses zero.
5. Read percentiles off the cash-out distribution. P10 = 10th percentile of the cash-out month (the month at which 10% of trials had run out). P50 = median. P90 = 90th percentile.

The output is a distribution, not a single number. The deterministic formula gives you something close to P50. The two new numbers, P10 and P90, are the decision-relevant statistics.

One nuance: gross burn is also volatile in real businesses, but typically with much lower CV than revenue (5 to 10% versus 30 to 50%). Holding burn constant and varying only revenue captures 80 to 90% of total cash-trajectory variance. Chunky annual cost events (insurance, software renewals, tax payments) should be modelled as scheduled outflows rather than averaged in.

4. Reading the runway distribution: P10, P50, P90

Three numbers, three different conversations.

P50 is your planning baseline. The median outcome. Half of futures are better, half worse. Decisions made assuming P50 happens are wrong roughly half the time, in either direction. Use it as the headline runway number when communicating to stakeholders who already understand variance.

P10 is your trigger-threshold number. The bad-case runway. In 10% of simulated futures, you run out of cash at or before this month. Anchoring triggers on P10 prevents the systematic late-trigger problem the deterministic formula creates. If P10 = 11 months, act as if you have 11 months of runway, not 15.

P90 is your upside ceiling. The good-case runway. Useful for one decision: how much of the upside you can spend down. If P90 says 21 months in a good scenario, that does not mean hire to 16 months of runway. It means optionality exists in good scenarios. Using P90 as a planning number is the founder equivalent of betting the household budget on a stretch sales target.

The gap between P10 and P50 is the cost of revenue volatility. A 4-month gap on a 15-month median, like the worked example below, is manageable. A 9-month gap (60%+ revenue CV) is a different category of risk and means the business needs revenue stabilisation as a higher priority than growth.

5. Trigger thresholds anchored on P10, not P50

Trigger thresholds are the action levels that force movement before the cash cliff. Standard practice (per the runway-planning piece) sets them at 18 / 12 / 9 / 6 / 4 months. Under revenue volatility, those thresholds apply to P10, not the deterministic average.

• P10 ≤ 18 months: Begin fundraise conversations or revenue stabilisation initiatives. For solo founders not raising, this is the threshold to start the next product launch or pricing change. Carta's Q4 2024 data shows median time between rounds in the 24 to 28 month range for venture companies^[1], making 18 months of P10 runway the practical floor for any company expecting to raise again.
• P10 ≤ 12 months: Discretionary-spend freeze. SaaS subscriptions audit. Vendor consolidation. The cost of doing this at month 12 is small; doing it at month 6 is operationally disruptive.
• P10 ≤ 9 months: Hiring freeze. Negotiate longer payment terms with major vendors. For solo founders, this is the threshold to defer new investment in tools, education, or contractor spend.
• P10 ≤ 6 months: Revenue-recovery work becomes the dominant priority. Bridge financing conversations, where applicable, start now because approval timelines run multi-month.
• P10 ≤ 4 months: Cash-preservation mode. Below this, the business is one bad month from existential risk regardless of scenario.

Founders who anchor triggers on the deterministic mean arrive at each threshold 2 to 4 months later than they should under typical volatility. Anchoring on P10 corrects the bias and is the single most consequential change in the honest method.

6. Worked example: 12-month projection at 30% revenue CV

A solo founder runs a productised consulting business. Cash on hand: $600,000. Gross monthly burn (their salary, contractor costs, software, taxes set aside): $100,000. Trailing-12 revenue averages $60,000/month with a standard deviation of $18,000 (CV = 30%). They want to know how much runway they actually have.

Inputs
  Starting cash               $600,000
  Gross monthly burn          $100,000
  Mean monthly revenue         $60,000
  Revenue std deviation        $18,000  (CV 30%)
  Distribution                 lognormal, monthly, independent
  Trials                       10,000
  Horizon                      36 months

Deterministic (single-point) runway
  Net burn                    $100k - $60k = $40k/month
  Runway                      $600k / $40k = 15.0 months

Simulation results (cash-out month distribution)
  P10                         11 months   (worst-decile case)
  P25                         13 months
  P50                         15 months   (median, matches deterministic)
  P75                         18 months
  P90                         21 months   (best-decile case)

  Probability of cash-out before month 12     19%
  Probability of cash-out before month 15     50%
  Probability of cash-out before month 18     78%

Trigger interpretation
  Anchor on P10 = 11 months
  18-month threshold       breached at month 0     start action now
  12-month threshold       breached at month 0     freeze active
  9-month threshold        breached at month 2     hiring/scope freeze
  6-month threshold        breached at month 5     revenue priority
  4-month threshold        breached at month 7     cash-preservation

The deterministic 15-month number is correct as a median statistic and useless as a planning anchor. Anyone planning against it is planning for a coin-flip outcome. The simulation reveals that a "comfortable" 15-month runway under 30% revenue CV has a 19% probability of cash-out inside 12 months. One in five futures runs out before the deterministic calculator predicts the company would even hit the first action threshold.

Action profile: anchor decisions on P10 = 11 months, treat 30% CV as the volatility-reduction target (cutting to 20% via revenue stabilisation moves P10 to roughly 13 months), and re-run the simulation monthly as new revenue data arrives. Once the spreadsheet exists, recurring work is about 30 minutes per month.

7. Common mistakes that hide volatility

Three patterns recur in mis-calculated runways. Each quietly converts a volatile business into a falsely stable one on paper.

Treating ARR as cash. Annual recurring revenue is a contract-value statistic, not a cash statistic. A SaaS with $1.2M ARR billed monthly collects $100k/month. The same $1.2M ARR billed annually with most contracts renewing in Q1 means cash collection is heavily front-loaded and Q4 cash sits substantially below the ARR-implied flat rate. Runway calculations done against ARR/12 instead of actual cash collection overstate cash availability in the back half of the year. FASB Topic 606 makes this distinction explicit on the accounting side^[4]; runway calculations should respect it on the cash side.

Ignoring deferred revenue. A SaaS with a strong sales motion can hold substantial deferred revenue on the balance sheet (cash collected for services not yet delivered). That cash is on hand but not freely available; it represents an obligation to deliver. A startup that collects $400k of annual prepayments in January has $400k more cash than monthly revenue would suggest, and $400k of forward cost-of-delivery booked. Net out the deferred-revenue balance, or run two runway numbers (cash-runway and obligation-adjusted runway).

Assuming constant CAC. Customer acquisition cost is a spend lever inside burn that scales with growth ambition, not a fixed line item. A business that holds CAC constant implicitly assumes growth is free, which contradicts the typical small-business pattern. The honest model treats CAC as a function of new-customer count and feeds the resulting marketing spend into burn. Our CAC calculator handles the input side; the simulation handles the variability.

A fourth mistake to flag: confusing accounting profitability with cash break-even, which our break-even units calculator separates into accounting and cash variants.

8. Where the calculators fit

The honest method does not replace the standard runway calculators; it sits in front of them.

• Startup runway calculator for the deterministic baseline. Run this first to get the P50-equivalent number. Treat its output as the median, not the answer.
• Burn multiple calculator to read efficiency. Burn multiple of 1.0 or below is the threshold OpenView's 2024 benchmarks identify with top-quartile capital efficiency^[3]; above 2.0 with high revenue volatility is the dangerous combination.
• Cash conversion cycle calculator for the working-capital component of volatility. Long collection cycles amplify revenue CV into cash CV; shortening the cycle is one of the few free reductions in cash variance.
• Break-even units calculator to distinguish accounting break-even from cash break-even. Useful when the question is whether the business needs more units sold or more dollars per unit to stabilise.
• CAC calculator to size the spend side of growth. The relationship between CAC, payback period, and runway is where founders most often miscalibrate.

The honest runway method is one piece of disciplined cash management, alongside cash-flow practice and three-statement projections. The point is the conversation it forces: how volatile is your revenue actually, and which of P10 or P50 is driving decisions today. As of 2026-Q2, with median time between funding rounds still extended^[1] and small-business revenue variability elevated relative to pre-2020 baselines^[2][5], the cost of running the deterministic version alone has gone up. The honest version takes thirty minutes a month and corrects a systematic error that compounds quietly until the day it does not.