Zoom: A Real Inflection, a Flat Guide, and a November Test

Zoom just produced the sort of quarter its turnaround story has been promising for two years — the strongest enterprise growth in three years, with paid AI sold into most of its biggest deals — and the stock fell anyway. The market was not objecting to the quarter. It was objecting to the plan for the next one: guidance that refuses to compound any of the improvement. Whether that forecast is caution or knowledge is now the entire question, and the answer shows up with the next earnings report in late November.

The quarter behind it

For the three months ended July 31, Zoom reported revenue of $1.277 billion, up 4.9% year over year and $7 million above the top of its own guidance. The engine inside that number was Enterprise: revenue of $787.5 million, up 7.8% — the strongest rate in three years — while the consumer and small-business Online line inched up just 0.6%. Larger customers already drive roughly 62% of total revenue, and the mix keeps shifting their way.

The AI story, long a narrative without a number, finally showed up in the deal flow. Paid AI features were included in nine of the top ten customer-experience deals of the quarter. Zoom Virtual Agent customers grew more than 250% year over year. Licensed monthly active users of Workplace AI features grew 125%. And the forward indicator that matters most — remaining performance obligations, or revenue already contracted but not yet recognized — rose 14% to about $4.5 billion, with the long-dated portion up 25%. That is the signature of enterprises signing bigger, multi-year platform contracts, and it is the first time in a while that the bookings data has supported what management keeps telling shareholders.

Why the market sold

Yet underneath the headline beat, the economics were soft, and the guidance delivered even less. Adjusted (non-GAAP) earnings of $1.55 a share were flat against $1.53 a year earlier and only about 5% above the $1.48 consensus. The GAAP operating margin fell to 24.6% from 26.4%, and the adjusted gross margin slipped to 79.1% from 79.8% — partly, management says, AI compute costs and the margin-dilutive Common Room acquisition. Free cash flow came in at $472 million, down from $508 million a year ago. And the customer base still is not growing on its own: enterprise net dollar retention, the measure of whether existing customers spend more or less over time, sits at 99% — up from 98% a year ago but still below the 100% line, meaning new-logo wins are offsetting churn rather than building on top of it.

The guidance is where the market found its reason to sell. Zoom guided fiscal third-quarter revenue to $1.275–1.280 billion, essentially flat against the quarter it just reported, and implied a slower growth rate than the 4.9% it just delivered. For the full year it raised revenue guidance to $5.085–5.095 billion — a midpoint barely higher than the $5.080–5.090 billion it set three months ago — despite beating the top of its own second-quarter range by $7 million. Full-year adjusted EPS guidance moved to $6.08–$6.12 and free cash flow to $1.78–$1.82 billion, but third-quarter EPS is guided to $1.46–$1.48, below the $1.55 just reported. Management also flagged a 40-to-60-basis-point top-line headwind from one churned white-label customer. Read plainly: after years of over-promising, management banked the beat rather than passing it through to the year. That is either conservatism on genuinely improving bookings — or a quiet admission that the growth pause continues.

What the price asks

The numbers explain the price action. The shares had run up roughly 29% in the four months before the report; they then fell about 7% from just over $100 into the mid-$90s before rebounding to about $100 in the following session. That leaves a market value near $29 billion against roughly $7.2 billion in cash and marketable securities, and the stock trading near 16 times forward adjusted earnings and about 18 times EV/EBITDA, with adjusted operating margins near 40% and a roughly 38% free-cash-flow margin. None of that is a distressed price. For a company guiding to roughly 4.5% top-line growth on a customer base that still shrinks a little on its own, it is a full one — a multiple being paid on the promise that the improving bookings eventually flow to revenue. The RPO data is the best evidence that promise is real; the flat guidance is the best evidence that it is not proven.

The November test

So the decision is a calendar one, and it is falsifiable in a single report. If Zoom beats the top of its third-quarter range and guides the fourth quarter to re-accelerating growth with enterprise still expanding above 8%, the "conservative guide" reading wins, the RPO book is validated, and 16 times forward earnings starts to look reasonable for a business that is compounding. If the third quarter lands in line, growth stays in the 4–5% band, retention stalls at or below 100%, and margins keep paying for AI, then the quarter's promise remains a promise and the multiple has room to compress.

The strongest bear fact against the AI-inflection case is not churn or the shrinking online business. It is that management, holding the best bookings evidence in three years, still would not guide to growth. That gap is either the opportunity or the warning — and the November print is where it gets resolved. Until then, the honest position is to watch, not to chase.