CrowdStrike's Guidance Now Demands Two More Record Quarters

CrowdStrike called its latest quarter the best in company history, and the headline numbers back it up. Net new annual recurring revenue — the sales engine of the business — hit a record $332.8 million in the quarter ended July 31, up 51% from a year ago and nearly $48 million, or about 17%, above the $284 million to $286 million it had guided. Ending ARR rose 25% to $5.84 billion. Total revenue came in at $1.47 billion, up 26%. After hours, the stock jumped about 9% to roughly $207.

The surprise wasn't the beat. The surprise was that the market rewarded it. CrowdStrike has spent 2026 running so far ahead of its own results that "beat" became the minimum payment rather than the reason to pay up for the shares. In June, after a first-quarter report that cleared both revenue and adjusted earnings, the stock fell 6.5% the next day. Ahead of this print, prediction markets put 89% odds on a ninth straight earnings beat, and the analyst board read 41 Buys against one Sell. That is a room with nobody left to disappoint — yet the stock went up anyway. The reason sits inside this specific quarter, and it's worth understanding before anyone chases the pop.

Start with what net new ARR means, because it is the number that leads this business. ARR is the annualized value of active subscription contracts; net new ARR is what the quarter added to that base. Each added dollar becomes roughly a dollar of subscription revenue over the following year, at a gross margin near 78%. Beating net new ARR guidance by 17% is not style points. It's compounding future revenue, not just a brighter quarter.

The acceleration is genuine in absolute terms — $332.8 million is the highest quarterly figure CrowdStrike has ever recorded, above the $330.7 million it posted in last year's seasonally strongest fourth quarter. One footnote on the "up 51%" pace, though: it measures against Q2 of fiscal 2026, a $221 million quarter that management itself was still framing as a recovery from the July 2024 outage. A low base flatters a percentage. The record itself stands on its own.

Then read the guidance the way you'd read a hand before betting. Management raised full-year net new ARR growth to a 34% midpoint, up 630 basis points from the 27.7% it committed to in June. Do the arithmetic CrowdStrike built into its own ending-ARR targets. Q1 and Q2 together banked $588.6 million of net new ARR. The raised outlook implies roughly $1.35 billion for the full fiscal year, taking ending ARR from $5.84 billion today to about $6.61 billion. The back half must therefore add roughly $765 million — which management's guidance splits into about $344 million implied for the third quarter and about $421 million in the fourth. Both would be records, and the fourth-quarter figure is about 27% above the record fourth quarter the company posted last January.

CrowdStrike has never delivered two quarters like that back to back. The company just made them the base case. That is not a stingy manager sandbagging the guide — this team has raised twice in a row and beaten its own bars on net new ARR. It is simply what the price demands.

At Wednesday's close of $189.18 — all per-share figures reflect CrowdStrike's four-for-one split in July — the stock trades at about 150x the raised fiscal 2027 adjusted EPS guidance of $1.25–$1.26, and roughly 32x forward revenue. The premium sits atop an expensive corner of the market: at about 38x trailing sales, CrowdStrike costs more than Palo Alto Networks at 26x, Fortinet at 15x, and Zscaler at 9x. Note how little the raise moved the denominators either — full-year adjusted EPS went to $1.25–$1.26 from $1.22–$1.24, about 2%. At 150x, even that small increment shifts billions of dollars of market value. That is the whole problem in miniature: the multiple amplifies every modest change into a large swing.

The longer view does not rescue it. One reasonable estimate: even if CrowdStrike compounds revenue 25% a year through fiscal 2032 and free cash flow margins reach 35%, today's enterprise value still works out to roughly 29x fiscal-2032 free cash flow. The market is not buying the next two quarters. It is buying a decade of flawless execution, at today's prices.

None of this is a criticism of the company, which passed every test the quarter offered. Revenue grew 26%. Gross and net retention both improved from an already-elite level near 97% and 115% in the spring quarter, and the quarter was profitable on a GAAP basis with free cash flow at about 26% of revenue. Falcon Flex, the consumption model pulling customers onto platform-wide spend, now represents $2.29 billion of ending ARR, up 101%, and AI-detection ARR more than tripled sequentially in the spring, with OpenAI, Anthropic, and Google signed on as coalition partners. Quality gate: clear. Moat gate: clear.

The valuation gate is the one that fails, and a disciplined framework has to say so plainly: the market is not wrong about CrowdStrike. It priced the reacceleration in advance — the shares began 2026 near $117 and hit an all-time high of $227.50 in mid-August — and it was right after the report, which is why the stock popped. There is no contrarian edge here, because the consensus is right. The risk now sits entirely on the delivery side.

So the discipline is to stop grading the next report on whether revenue "beats." Grade it on whether third-quarter net new ARR clears the roughly $344 million the guidance implies and whether the back half makes good on the roughly $765 million bridge. A quarter that merely small-beats revenue while net new ARR undershoots that line is a re-rating event at 150x — the multiple is the downside, and it has more room to fall than the stock has to rise off a beat that is already in the price. A quarter that clears the bridge is exactly what the market is already paying for. Neither outcome favors chasing a record quarter at 150x earnings. The better play is to let the second half deliver the bridge, or to wait for an entry that does not demand two of the steepest quarters in CrowdStrike's history as your margin of safety. Either way, the number that decides this stock is not in the report you just read. It is in the two reports you have not.