The Rule of 40 broke this quarter, and three names didn't tell you.
Median Rule of 40 across the 144-name bench rose 360 basis points to 38.4 — the largest single-quarter expansion in five years. The headline is real. The composition is misleading.
The bench moved up because the AI-native infrastructure cohort delivered an outsized contribution to the median, while a quieter set of legacy app SaaS names compressed underneath it. If we strip the top decile out of the cohort, median Rule of 40 falls back to 31.6, which is roughly where it sat at the prior reading. The expansion, in other words, is a story about thirteen names, not one hundred and forty-four.
Three of the names contributing to that expansion are doing so on a basis that the disclosure does not clarify. We surface them below.
01What the bench median says
The headline number is the median across the covered cohort. We compute Rule of 40 as trailing-four-quarter ARR YoY plus trailing-four-quarter free-cash-flow margin, on a constant-cohort basis. The calculation is reproduced in our methodology page; the inputs are filed-quarter financials, with one-time items normalised on a published basis.
Figure 1
What the median misses is the dispersion underneath it. We split the cohort into the top and bottom halves and find the gap between the two has widened from 12.4 points one year ago to 18.8 points this quarter — the widest dispersion we have measured.
02Where the expansion came from
Roughly 70% of the bench-wide expansion in Rule of 40 is attributable to thirteen names sitting in the AI-native infrastructure cohort. These names benefitted from a combination of (i) high incremental gross margin on consumption revenue, (ii) S&M leverage from inbound demand, and (iii) a step-function in net new ARR.
The remaining 30% of the expansion is distributed thinly across the cohort. We did not find a second cluster.
03The three names that didn't tell you
Three names in the top contributor set are achieving the expansion on a basis that, in our reading of their disclosures, materially overstates the comparable figure.
First. One large-cap data-infrastructure name disclosed a Rule of 40 of 52.4. We compute, on the SaaSO standard with consumption credits normalised, a comparable figure of 46.1. The 6.3-point gap is attributable to an upfront recognition of multi-year prepaid AI-credit balances that we do not regard as durable ARR.
Second. A SecOps name reports a Rule of 40 of 56.1. The disclosure includes a one-time deferred-revenue release that we normalise out, bringing our comparable figure to 49.7. We are still comfortable with the underlying business; we are not comfortable with the headline.
Third. An observability name reports a Rule of 40 of 53.4 and an FCF margin of 28.4%. The cash-flow figure includes a 110 bps benefit from a working-capital tailwind that, on our normalised cycle, will not repeat. The comparable forward figure is closer to 51.0.
04What the bench looks like, normalised
If we apply the SaaSO normalisations across the bench, the median Rule of 40 falls from the disclosed 38.4 to a comparable 35.2 — still a 240-bps expansion year-over-year, but a meaningfully smaller one.
Figure 2
05What we'd do
We continue to favour the cohort on a normalised basis. Our preferred screen pairs (i) normalised Rule of 40 ≥ 40, (ii) NDR ≥ 110%, and (iii) magic number ≥ 0.75. Five of the 144 names clear all three. Two of those five sit outside the visible top contributor set.
We will publish those two names in next week's note (RPT-219), with a one-page screen.