MongoDB FY 2Q27: A clean Beat and Raise!

1. The Headline

MongoDB delivered what management repeatedly called a “very strong” quarter that beat every guidance range, prompting a second-half raise driven primarily by Atlas.

Metric	Q2 FY27	Comparison
Total revenue	$771.8M	+30% YoY (vs. 24% a year ago) — highest growth since FY24
Atlas revenue	~$2.3B run-rate	~+29% YoY (5th straight quarter at ~29%); beat guide by ~300 bps
EA & other revenue	—	~+30% YoY (strongest in 3 years); reported as +36% in CEO remarks
Net ARR expansion rate	122%	vs. 119% a year ago, 121% last quarter
GAAP operating income	$28.4M	We’re positevely surprised by GAAP income
Operating cash flow / FCF	$142M / $139M	roughly doubled from $72M / $71M a year ago
Customers	70,600	+2,900 net new (a record); 69,300 on Atlas
$100k+ ARR customers	~3,000	+17% YoY
RPO	$1.52B	+91% YoY (current portion +73%)

The core story: two growth engines now working simultaneously — cloud (Atlas) and self-managed (Enterprise Advanced) — with an emerging, still-small AI contribution layered on top.

Revenue Acceleration: 30% YoY!

In MongoDB case we can clearly see the scale doing its job both on Gross Margin, but also on GAAP Operating Income.

GAAP Net Income: Positive for the 3rd consecutive quarter;

MongoDB generated $139 million in Free Cash Flow in 2Q26 (FY 2Q27) and $338 million so far this year!

Against our numbers:

We were impressed with MongoDB acceleration in 2Q26 which is proving the company will be a main beneficiary from AI.

Also the scale helped improve margins beyond what we expected, mainly in Cost of Revenue and Selling and G&A! MDB P&D shows the company is keeping investing in its products. But the scale benefit to margins were pretty cleat this quarter!

2. Key Themes

A. The strategic positioning

CJ Desai framed MongoDB as “emerging as the real-time intelligent data platform for modern applications in the multi-cloud and AI era.” The central argument: because enterprises already run large parts of their data estate on MongoDB, building an AI agent on top is a natural extension — “the data an agent actually needs is live operational data, not a stale copy sitting in a warehouse.” Search, Vector Search, and Embeddings are “built in, not bolted on.”

B. “Run anywhere” as the key differentiator

This was the quarter’s recurring phrase. The advantage is that customers can run MongoDB across public clouds, neo clouds, on-prem/colos, and self-managed environments — and management insists it is an “and, not an or” between Atlas and EA.

C. Three AI demand vectors

1. Enterprises (Global 2000 / Fortune 500) — adding agents on existing data estates.
   • Financial Times: unified operational data + Vector Embeddings on one platform; hybrid full-text + semantic search; uses Voyage 4 for indexing and the cheaper voyage-4-lite to serve 100,000+ daily queries, cutting retrieval cost. Index work that “took weeks” now finishes “in a day.”
2. Frontier labs (both customers and partners):
   • One lab migrated its chat memory system to Atlas in 4 weeks after leaving Postgres (performance lags/outages) — now runs “10x faster reads than Postgres.”
   • Labs also use MongoDB for research workloads (experimental results, evaluation data, training artifacts).
   • MongoDB launched a fully managed MCP Server for Claude Code, Codex, Grok Build, Cursor, and Devin (Cognition) — keeping MongoDB “embedded in the AI supply chain.”
   • Anthropic’s Chief Commercial Officer Paul Smith endorsed the partnership, noting developers building on Claude are pointed to MongoDB Voyage for embeddings.
3. AI natives — some choose MongoDB from day 1, others migrate in as they hit scaling limits.
   • Fireflies (unicorn): chose Atlas from day 1; 20M+ users, 1M+ orgs, 7B+ meeting minutes, 40+ microservices.
   • Eve (unicorn, legal): uses Atlas + Voyage’s Rerank 2.5 for RAG over case documents.

D. Voyage momentum

• Voyage customer count nearly doubled QoQ (second consecutive quarter).
• A large majority of new Voyage customers are AI-native with no prior MongoDB relationship — a new top-of-funnel.
• New launches in August: Automated Voyage Embeddings in Atlas (one-click Vector Search setup), voyage-code-4 (code-purpose model), and an upgraded Reranking API.

E. Enterprise Advanced (EA) — the surprise strength

Two structural patterns explained EA’s standout quarter:

1. AI in governed, self-managed environments — Search and Vector Search were brought to EA (GA in July/”delivered June 30″), closing the gap with the cloud experience. Demand “came in immediately.” A major U.S. bank extended EA (already running 100+ production apps across payments/fraud) to GenAI and semantic search while keeping sensitive data inside its own governed environment.
2. Hybrid deployment — customers on both EA and Atlas simultaneously. Nationwide U.K. (world’s largest building society) runs a “speed layer” across both, supporting 24M+ weekly app logins, up to 3M transactions and £1.5B on a peak day. EA’s profitability funds R&D and go-to-market that further Atlas growth.

F. Leadership

CJ highlighted a refreshed team: Ben Cefalo (core products), Pablo Stern-Plaza (AI/emerging products), and new CRO Ryan Mac Ban.

3. Financial Detail (CFO Mike Berry)

• Atlas: 6th straight quarter of increasing YoY dollar growth, adding a record $127M in the quarter. Growth led by North America and the $100k+ ARR cohort (upmarket momentum). Of $100k+ Atlas customers, 48% use 2+ features (up from 42%), driven by Vector/text search.
• EA & other ARR (normalized for duration): ~+11% YoY — third consecutive quarter of double-digit ARR growth.
• Profitability: 3rd consecutive quarter of GAAP EPS profitability; full-year FY27 guided to be GAAP EPS profitable.
• Balance sheet: $2.4B cash/equivalents/short-term investments. $100M share repurchases in the quarter; $59M to settle RSU taxes. First-half FY27: $344M operating cash flow, $335M FCF.

Guidance (raised)

	Q3 FY27	Full-Year FY27
Total revenue	$756M–$761M (+20–21%)	$2.99B–$3.03B (+21–23%)
Atlas growth	~26%	~27% (raised +300 bps from prior midpoint)
EA & other	mid-single-digit growth; ~flat H2	~11% (up from mid-single-digit) — first double-digit full year in 3 years
Non-GAAP op. income	$152M–$156M (~20.5% margin at high end)	$616.3M–$636.3M (~21% margin at high end)
Non-GAAP EPS	$1.57–$1.61 (87.1M shares)	$6.39–$6.58 (86.4M shares)

• Operating margin expansion now ~250 bps for FY27 (100 bps above prior high end).
• FCF conversion: now expected at the upper end of the 80–100% long-term target.
• “Rule of 44” targeted at the high end (23% growth + 21% op margin).
• Q3 is flagged as the toughest Atlas compare of the year. Guidance framework unchanged: ~200–300 bps of outperformance vs. initial guide over the last 3 quarters.

3. Key Takeaways

What genuinely changed this quarter:

• Acceleration is real and broad. 30% total growth is the fastest since FY24, and it came from two engines, not one. EA’s ~+30% (strongest in 3 years) was the biggest positive surprise, and it’s structural — the June 30 launch of Search/Vector Search on EA created net-new, monetized demand (they charge extra) that management didn’t previously model.
• The AI contribution is still small but has crossed from anecdote into signals. Voyage customers doubling QoQ (twice), Vector Search outpacing company growth, rising MCP-connected clusters, and multi-feature adoption (48% vs. 42%) are the leading indicators. Actual AI revenue remains “small” by management’s own admission.
• The Postgres displacement narrative got concrete — a frontier lab citing “10x faster reads” and a 4-week migration is the sharpest competitive claim on the call.

What management deliberately hedged:

• Q4 Atlas deceleration is conservatism, not weakness. Two things de-risk the bear case: (1) net ARR expansion rose to 122%, and (2) Berry volunteered that there were no large bundled deals this quarter — pre-empting the “was this a repeat of last Q4’s EA→Atlas flip?” concern. The implied Q4 slowdown is the standard guidance framework applied to the toughest compare of the year.
• EA guidance is intentionally cautious because of multiyear-deal lumpiness — management explicitly invited investors to judge EA on full-year rather than quarterly figures, and hinted at upside if deals land larger.

The strategic tension to watch:

• The single most important recurring message — repeated by both executives across at least four questions — is that EA is not cannibalizing Atlas. This defensiveness signals it’s the key investor worry. The mix answer (Atlas ~74%, rising but more slowly) confirms EA is now taking a larger share of the growth than originally planned. The bull framing is “two durable growth drivers”; the bear framing would be that the higher-margin, faster-growing cloud transition is decelerating relative to prior expectations. Management is firmly in the “and, not or” camp.
• The Voyage top-of-funnel is a genuine option value — free/cheap AI-native customer acquisition driven by coding-agent referrals (Claude, Codex), with cross-sell to Atlas still almost entirely ahead of them. The awareness gap (“we didn’t know Voyage is MongoDB”) is both a marketing failure and an untapped conversion lever.

Quality of the print: Beat on every line, margin expansion, cash flow doubling, GAAP profitability for a third straight quarter, RPO +91%, a record 2,900 net-new customers, and a Rule-of-44 full-year target. This is a clean, high-quality beat-and-raise with the added narrative upgrade of EA re-rating from a mature/declining line into a double-digit grower.

This is an analytical piece only. For Opinions, Valuation and Stock Expected Return, please subscribe!

Thank you!

CoreValue Research Team