CRWD · Forward model · Subscription · Bear case
What has to happen in Subscription
Model as of
This page changes Subscription inside the complete CRWD model. Every other vertical, corporate programme and valuation assumption stays in the selected portfolio case.
Shares this vertical and portfolio case. Slider and horizon edits stay in your browser.
Subscription
Basis quarter$1.40B
Final quarter$2.19B
Implied CAGR+9%
Final revenue mix97%
The Falcon platform sold as recurring subscriptions, and 95% of revenue. What a quarter can earn is set by the annual recurring revenue already installed at the end of it: a dollar of ARR turns into a quarter of itself in a quarter. Growth is therefore net new ARR compounding on a base that renews at a 115% dollar-based net retention rate, which is why this line is modelled as installed capacity rather than as customers times a price CrowdStrike does not publish.
Last four quarters
2026 Q3
$1.17B
Reported
2026 Q4
$1.24B
Reported
2027 Q1
$1.32B
Reported
2027 Q2
$1.40B
Reported
Falcon platform modulesFalcon Flex commitmentsManaged services
ARR base
5841 $M of ARR
at the basis quarter
ARR of $5,841.4M at 31 July 2026. Derived, not printed: the release rounds to "$5.84 billion" and discloses $332.8M of net new ARR in the quarter, which on the $5,508.6M the Q1 10-Q disclosed gives $5,841.4M. The July-quarter 10-Q, which prints ARR to a tenth of a million, had not been filed when this was written.
Net new ARR
345 $M of ARR/qtr
changing +3.0% per quarter
$345.0M, the Q3 FY2027 ARR guidance midpoint of $6,186.4M less the $5,841.4M installed. Management's number for the next quarter's build, and it implies 29-31% year-over-year growth in net new ARR.
Utilisation
94%
share of a quarter of ending ARR that lands as revenue, gliding toward 97%
93.83%, solved so that one glide step (to 94.02%) applied to guided Q3 ending ARR reproduces the Q3 revenue guide midpoint. It is well below the 95.89% the basis quarter actually delivered, because the Q3 revenue guide is roughly 1.9 points of utilisation - about $29M - below what the same guide's ARR would produce at the realised ratio. Calibrating to the guide imports that conservatism into every quarter.
Revenue per $M of ARR
$250000/qtr
drifting 0.0% per quarter
Definitional: a dollar of annual recurring revenue yields $0.25 in a quarter, so $250,000 per $1M of ARR.
Subscription
Latest: $2.19B (2032Q2E)
| Period | Value |
|---|---|
| 2024Q2 | $690M |
| 2024Q3 | $733M |
| 2024Q4 | $796M |
| 2025Q1 | $872M |
| 2025Q2 | $918M |
| 2025Q3 | $963M |
| 2025Q4 | $1.01B |
| 2026Q1 | $1.05B |
| 2026Q2 | $1.10B |
| 2026Q3 | $1.17B |
| 2026Q4 | $1.24B |
| 2027Q1 | $1.32B |
| 2027Q2 | $1.40B |
| 2027Q3E | $1.42B |
| 2027Q4E | $1.46B |
| 2028Q1E | $1.51B |
| 2028Q2E | $1.56B |
| 2028Q3E | $1.60B |
| 2028Q4E | $1.64B |
| 2029Q1E | $1.69B |
| 2029Q2E | $1.73B |
| 2029Q3E | $1.77B |
| 2029Q4E | $1.81B |
| 2030Q1E | $1.85B |
| 2030Q2E | $1.89B |
| 2030Q3E | $1.93B |
| 2030Q4E | $1.97B |
| 2031Q1E | $2.01B |
| 2031Q2E | $2.04B |
| 2031Q3E | $2.08B |
| 2031Q4E | $2.12B |
| 2032Q1E | $2.15B |
| 2032Q2E | $2.19B |
Assumptions & reasoning
- Every quarter here is subscription revenue as reported on the face of a CrowdStrike release. Nothing is apportioned and nothing is estimated. The basis quarter is $1,400,291k, three months ended 31 July 2026, Exhibit 99.1 of the 8-K filed 26 August 2026.
- ARR is measured at period end while revenue accrues through the quarter, which is the whole job of the utilisation input: it was 95.89% of a quarter of ending ARR in the basis quarter, 95.91% the quarter before, and has sat between 94.6% and 95.9% for two years.
- The 115% dollar-based net retention rate in the thesis is no longer a published number. The release and the slide appendix both define net and gross retention and say both improved sequentially, but neither prints a percentage, and CrowdStrike has not resumed printing one. The model does not depend on the rate - it drives off ARR, which is disclosed - but the thesis quotes a figure the company will not confirm.
- NET NEW ARR ACCELERATED AND THE DRIVER DOES NOT. Net new ARR was a record $332.8M, up 51% year over year, and the FY2027 net new ARR guide was raised 630bp to +34% growth after a 520bp raise a quarter earlier. The driver still compounds net new ARR at 3% a quarter, 12.6% a year. That assumption is now the loosest thing in this model and a future revision has to settle it; it was left alone here because this pass moves the basis, not the argument.
- CrowdStrike allocates no operating expense, depreciation or capex to the revenue split, so the margin and capex inputs on this line are ours and no disclosure tests them. Capex quadrupled year over year in the basis quarter, from $30.5M to $124.4M, or 8.5% of revenue against the 6.5% intensity carried here - another thing left alone.
- Net new ARR is heavily fourth-quarter weighted and this projection carries no seasonality, so a constant build rate lands fiscal 2027 exit ARR at $6,542M, about 1% below the guided $6,603.0-6,611.9M range, while hitting the revenue guide.
- The previous basis projected $1,375.0M of subscription revenue for this quarter against $1,400.3M reported, a 1.8% miss, all of it the guide-calibrated utilisation being too low.