7 SaaS Metrics That Separate Winners From Laggards: What CrowdStrike's 270% vs SentinelOne's -63% Reveals

In early 2025 Anthropic and OpenAI rolled out tool after tool and the market priced a broad software scare, the so-called SaaS-pocalypse. Valuations slid as if every subscription business faced the same threat. Time showed the opposite for some names: a subset not only survived but gathered strength under pressure. The video asks how to spot that split in advance and why looking only at revenue can mislead.

What Actually Recurs? — ARR and Net New ARR

The core of subscription health is not services or one-off licenses but the subscription itself. Annual recurring revenue takes the monthly subscription value of the active base and multiplies by 12, so new customers within the year are captured immediately; trailing revenue looks backward instead. Net new ARR, the period-to-period change, reveals true momentum. Thresholds are clear: above 30% is hyper-growth, 15-30% is solid, below 15% is not a high-octane SaaS profile. CrowdStrike reaccelerated net ARR growth from 20% in July 2025 to 25% and added about $300 million of net new ARR in the latest quarter, crossing $5 billion to $5.25 billion ending ARR. SentinelOne stayed flat at around 20% on a far smaller base with no reacceleration and a drifting trend, showing how growth quality diverges even inside the same cybersecurity cohort. ARR is not audited, so each company's footnote on how it is computed must be read.

What Does a Dollar Cost? — Gross Margin Direction

Subscription gross margin should be very high because serving one more customer costs little, so 80% plus is the ideal. Many firms only disclose total gross margin, not the subscription cut; then direction matters as much as level. AI features added a new cost line; code used to be write-and-ship with little ongoing burden, now inference has a direct cost. CrowdStrike expanded total gross margin from about 74% in January 2025 to 75-76%, so AI costs did not erode profitability and were offset by pricing or attach. SentinelOne moved the other way, from 75% in mid-2024 to 72%; the slide is not dramatic but signals weaker pricing power and a narrower moat when both faced the same cost shock. Rising margin tells you how much of each dollar stays and whether scale economics work.

Does the customer base spend more over time? Dollar-based net retention tracks the same cohort a year apart, dividing this year's spend by last year's, including those who churned. Above 100% means you keep the base, above 120% is gold standard, 105-120% is still healthy, below 105% is a warning. Every firm defines the cohort differently, so the footnote matters more than the headline. CrowdStrike last disclosed 115% and said on the last two calls it rose sequentially, implying 15-20% expansion just from existing customers. SentinelOne leaves a hazier picture: the last calculation covers early 2025 with almost a two-year gap and the definition was narrowed to customers spending at least $100,000. The transparency gap itself is a signal; one shows a rising spend curve, the other reduces visibility.

What Hides Behind Churn? — Gross Retention

Net retention can mask churn because new logos flatter the number. Gross retention caps each cohort member at last year's spend and isolates attrition. At least 90% is wanted, 95% plus is strong; avoiding a full rebuild of the base each year eases the cash cycle. CrowdStrike discloses gross retention consistently at 97-98%, meaning only about 3% of customer dollars churn each year, and management said it kept improving in the last two quarters. SentinelOne does not disclose the figure; absence is not neutral because the investor cannot model base erosion. Reading net and gross together separates growth quality from fragility.

How much is already sold answers forward visibility and is captured by remaining performance obligations. Think of a state fair where cash is exchanged for tickets at the booth and only counted as revenue when rides are taken; RPO is the stack of tickets already paid at the booth. Compare its growth to revenue growth: faster means the pipeline is strengthening, similar is balanced, slower warns of deceleration. Contract length can distort the level, so tenor must be considered. CrowdStrike RPO growth jumped from 29% to 49% while revenue grew 26% in the same quarter, with total RPO at $10.7 billion. SentinelOne shows a similar pattern on a smaller scale: RPO up 42% versus revenue 21%, good momentum but flattered by base size.

The 12-Month Lead — Current RPO

What share converts within the next year is current RPO and it leads revenue. It is total RPO times the current portion. At CrowdStrike 46% of $10.7 billion is current versus 52% of $7.2 billion a year ago, implying about 49% current RPO growth and a clear lead over revenue. SentinelOne reports on a 24-month horizon instead of 12, so the math uses 79% of $1.7 billion versus 85% of $1.2 billion and yields about 32% growth, 17 points behind. The horizon mismatch blurs direct comparison and should be normalized before judgment. Still, in both cases current growth outpacing revenue suggests near-term support.

Finally does any cash come out? Net income uses accrual accounting; cash generation needs cash from operations minus capital expenditures and internal software costs divided by revenue. The goal is 25% plus free cash flow margin, with quarterly wobble around investment cycles. CrowdStrike has long been positive, typically high-20s to low-30s and 25-36% over the last twelve months, keeping more than 25 cents of every dollar. SentinelOne is the opposite: occasional positive quarters but broadly negative free cash flow margin, so scale and pricing have not yet turned into cash. A cash-generative SaaS can invest through a downturn; a cash-consuming one needs external funding.

Together these seven forward signals explain why revenue alone lied. CrowdStrike delivered about 270% over five years while SentinelOne fell about 63%; at one point they looked similar and the forward metrics split them. The video walks through Stock Simplified Pro, powered by Fiscal AI, under financials → quarterly → key metrics to pull ARR, RPO and margin curves and layers moat, growth and valuation scores. The lesson is not to crown one metric but to look for direction and consistency together. A cheap multiple alone is not a buy case; when accelerating ARR, expanding margin, net retention above 105%, gross retention near 97%, RPO and current RPO growth ahead of revenue and cash margin above 25% point the same way, the discount becomes meaningful. The tool speeds the checklist but judgment stays with the investor; footnotes and call commentary should be rechecked each quarter.