Those band medians are medians of different companies, not one company priced at two growth rates, and that distinction is where the whole argument either holds or quietly falls apart: the file says the 20-to-30% cohort trades at 77% more revenue multiple than the 10-to-20% cohort [6], but it does not say the market would hand any particular business that step for one quarter of acceleration, since the faster cohort is also a different set of companies with different retention and different cost structures.
The instrument is also coarser than the gaps it is being used to adjudicate. SaaStr annualizes by multiplying the reported quarter by four and says outright that seasonality will move the figure by a few points [9], and then Atlassian, the second-largest name on either list at roughly $7B annualized, misses the line by two [6]. On that base, two points is about $140M of additional annual revenue [4]. Snowflake cleared the line at $5.6B while CrowdStrike missed it at $5.5B [7], which is $100M of scale between them [5] and, on SaaStr's reading, a line that separates one valuation band from another [16].
Four of the seven collect the AI budget without a sales cycle, because Palantir, Datadog, Cloudflare and Snowflake bill against usage and the invoice rises when the customer's workloads do [10]. Rubrik and Samsara do not fit that pattern and sell into structural change instead, data security and recovery for AI deployments in one case, digitizing physical operations in the other [11]. Figma is the copyable one: the most seat-dependent business on the list grew 48% in the quarter ended June 30 with net dollar retention at 136% [12], having added an AI credit layer on top of seats rather than in place of them, with roughly two-thirds of customers above $10,000 in ARR taking full seats at renewal and one large technology customer adding more than 25,000 paid seats through the credit add-on [13]. The bill showed up in gross margin, five points lower year over year, because products in beta carry inference cost and charge nothing for it [14].
This is probably wrong, but the metric worth writing into a plan is the share of revenue that grows without a renegotiation, which is the one thing the seven share across two very different pricing models [10][11]. The counter-thesis is timing. If the AI spending that lifted these seven eventually reaches the 41 of 58 index constituents currently growing under 20% [2], then 18% is a trough rather than a benchmark, and any board that re-bases targets to it now has re-based at the bottom [3]. Figma's own guidance cuts both ways, implying 36% for the September quarter against 48% delivered, a twelve-point step down that still clears the line by six [15][7]. The test is the 11 companies sitting in the 20-to-30% band [4]: if most of them cross up over the next four quarters, the median is moving and this reset dates from 2025; if most drift down, 30% stays a top-decile outcome [1] and a handful of growth points keeps costing more than a decade of comparables would suggest [16].
