MongoDB Delivered Its Strongest Growth in Years — and the Market Punished It

On September 1, MongoDB reported the sort of quarter software investors usually cheer. Revenue rose 30% year over year to $772 million, its fastest growth since fiscal 2024, and the company raised full-year guidance. The stock promptly fell more than 14% and now sits roughly 22% below where it traded 20 trading days ago.

When a company's best quarter in years gets sold, the market is not questioning the level of growth. It is questioning the quality of that growth — where it came from and whether it can last. MongoDB's Investor Day in New York on September 29 is where management gets to answer that question, under a CEO who only took the job last fall.

A growth quarter the market didn't trust

The secret is in the split of that 30%. MongoDB's cloud database, Atlas, is 73% of revenue and the engine investors actually value. Atlas grew 29% year over year — the fifth consecutive quarter at roughly that pace, and a record $127 million in added revenue. But Atlas is not what accelerated. The acceleration came from Enterprise Advanced, the on-premises and licensed line, which grew 36%. Management itself calls the segment "lumpy" because multi-year on-premises and hybrid deals are recognized in full up front, and regulated industries bought unusually large commitments in the quarter.

That is the whole bear case in three lines: a headline that accelerates on one-time license timing tells you nothing about whether the underlying cloud business is re-accelerating. Five straight quarters of Atlas stuck near 29% is the fact investors latched onto, and the reason a beat and a guidance raise turned into a 14% selloff.

Management was honest about the lumpiness, which is its own confirmation of the worry. The third-quarter guide implies revenue falls sequentially to roughly $756–761 million from $771.8 million, with EPS stepping down to about $1.57–1.61 from the just-reported $1.90, on tough comparisons and that same on-premises accounting volatility. The full-year guidance raise — to $2.99–3.03 billion from $2.92–2.96 billion — largely captures the quarterly beat and assumes a stronger second half.

None of this makes the company weak. Free cash flow nearly doubled to $137.6 million, non-GAAP operating margin reached about 24%, and remaining performance obligations accelerated 91% to $1.52 billion — signed future commitments are real and growing. Customers passed 70,600, up 2,900 in the quarter. The Voyage AI assets management bought last year pressed higher, with customer counts nearly doubling sequentially, though AI is not yet a material revenue driver.

The multiple reset, and the moment that matters

After the slide, MongoDB trades near $368, around 10.7 times trailing revenue against roughly 20 times for Datadog and about 39 times for CrowdStrike. In the language of this setup: the valuation has reset faster than the business has deteriorated. Over the trailing year the stock has traded as high as $473, and near its March low it had crashed 26% on a weak cloud-outlook quarter. This is a stock whose multiple swings hard on whichever way the cloud narrative tilts.

Cheap relative to richer peers only helps if the engine investors pay for — Atlas — shows it can re-accelerate. That is precisely what the September 29 Investor Day is positioned to test. It is the first major forum for CJ Desai, the former ServiceNow executive who became CEO on November 10, taking over from Dev Ittycheria after an 11-year run. A new CEO's first investor day is the natural place to set long-term targets and to turn the AI story — vector search, Voyage, the "agentic era" framing — from pitch into numbers.

Good company, unresolved stock

Separate the two. This is a good company: real growth, raised guidance, robust cash generation, and a balance sheet that needs no hand-wringing. The stock question is narrower and not yet answered — is 30% durable, or is it flattered by on-premises timing that reverses next quarter? Five quarters of Atlas at ~29% is the strongest bear fact, and no reported number has overturned it yet.

The reset has done real work. The bar to be constructive at $368 is much lower than it was near $473, and the investor day plus the next two quarters form a clean proof window. But until Desai either shows Atlas re-accelerating or hands investors numbers that reframe the growth, "too early" is the more honest read than a forced buy. Watch what the new CEO says about Atlas and how much of fiscal 2028's growth he is willing to commit to — that is the test that turns this selloff into either a bargain or the latest lesson in trusting the cloud graph before the license line.