MongoDB (MDB)

Executive Summary & Action Plan

Verdict

HOLD — The best-behaved financials in the data cohort (Atlas +29%, NRR accelerating to 121%, first GAAP profits, SBC growing 4% against 25% revenue) are shadowed by the worst-behaved leading indicators: Postgres hit 55.6% developer usage while MongoDB was the only major database to shrink, and AI coding agents scaffold new apps on pgvector, not Atlas. At $342 the price splits the difference; so do we.

Price $342.08 | Market cap $27.5B | Target $305 | Upside -10.8% IWANNAVY Fair Value · price as of 2026-07-11 close · Street consensus $394.68 (reference; n=36, 69% dispersion — the tightest in our data cohort) — our gap comes from expensing SBC (~96% of plain FCF) and pricing the funnel-decay risk the Street's lagging-metric models don't see

Abstract

MongoDB's reported quarter was the cleanest in our data-platform sweep: Atlas grew 29.4% with a record $117M YoY dollar add, net ARR expansion accelerated to 121%, RPO jumped 88% on the two largest deals in company history (a $90M Atlas TCV and a >$100M Enterprise Advanced contract), GAAP net income turned positive, and FCF margin reached 29% — while SBC grew only 4% against 25% revenue growth, the best hygiene trajectory in the cohort. The fiscal Q4→Q1 revenue dip is ASC 606 license-timing seasonality, not demand. The bear case lives upstream of the financials: Stack Overflow's December survey put Postgres at 55.6% usage with MongoDB the only major database shrinking, and 60%+ of new Supabase databases are AI-tool-created — the agent-era greenfield default is being set on pgvector while Atlas's installed-base moat holds. Our $305 fair value is the log-median of an SBC-expensed DCF ($150), 28x forward FCF ($305), and growth-adjusted sales ($337). Rating: HOLD — the only data name we don't rate SELL today; add at $305/$270, trim above $365.

Action Plan

Reports LAST in the August gauntlet (DDOG 8/6 → PLTR 8/10 → SNOW 8/26 → MDB 8/27E) — three weeks of sympathy risk before it can defend its own numbers. Hold core; add $305/$270; trim $365+; stop for traders $255.

• Risk/Reward 0.9:1 at market, ~2.6:1 from $270 | Prob-weighted 12M return −9.4% (Bull 30% × +6.7% + Base 45% × −10.8% + Bear 25% × −26.5%) | Confidence: Medium

---

Last Four Quarters

Fiscal quarters (FYE January). Gross margin ~72%, GAAP operating margin marched from −11% to breakeven territory, and FCF margin stepped 12% → 29% in four quarters. The Q4→Q1 sequential revenue dip ($695M→$688M) is mechanical: fiscal Q4 concentrates multi-year EA license renewals recognized upfront, and fiscal Q1 has three fewer consumption days — Atlas itself grew ~$12M QoQ. Net profitability currently leans on interest income from the $2.4B cash pile; the operating line is the honest gauge.

GAP verification (−5.73% on the 7/10 session): no fresh negative catalyst — profit-taking and rotation out of high-multiple software on an up-market day; MDB was hit hardest as the biggest recent gainer, with post-run insider-sale filings digesting (Quiver, 7/10).

---

Revenue & Profit Mix

Subscription is 97% of revenue; Atlas is the engine (75%, +29.4%), EA the lumpy-but-durable annuity (+13% — the >$100M record deal was an EA contract). FY27 Q1 per the 5/28 8-K.

---

Business Lines

• Atlas (75% of revenue, +29.4%): Consumption DBaaS, multi-cloud, 66,400+ customers, ~$2B run-rate. Must know — revenue follows application usage, not bookings; the $90M multi-year TCV deal shows Atlas winning genuinely large enterprise commitments, not just long-tail workloads.
• Enterprise Advanced (~22%, +13%): Self-managed term licenses, fiscal-Q4-concentrated recognition. Must know — EA is proving unexpectedly durable (largest-ever TCV contract was EA; Search/Vector Search went GA for EA on 7/2) — regulated and sovereign workloads can't all move to cloud.
• Voyage AI / vector stack (<1% of revenue): Embeddings + reranking fused into Atlas; MCP server for agent access. Must know — Voyage customers more than doubled QoQ; the thesis is that embeddings + vectors + operational data in one database beats pgvector bolted onto Postgres. Leading-indicator, not P&L, evidence so far.
• Community funnel ($0 revenue): Must know — this is the contested front: if AI coding tools default new projects to Postgres, the funnel erodes silently for 2–3 years before reported growth shows it. Watch new-workload adds, not just expansion.

---

IWANNAVY Fair Value

Log-median lands at $305. The spread is the debate: the relative lens ($337) says MDB is the cheapest of an expensive cohort per growth point (0.39 vs SNOW 0.54, DDOG 0.85); the SBC-expensed DCF ($150) says owner earnings today are ~$19M against a $27.5B market cap. Even our bull case ($365, multiples path) sits only 7% above spot — a fair price for a good business with a contested funnel.

• Thesis breaker (toward SELL): Atlas printing a 27-handle or the FY27 guide merely reiterated on 8/27, or developer-survey/funnel data worsening another leg — confirms the guided H2 deceleration (~15%) is real demand decay, opening the $240–270 zone. (toward BUY): a first AI-workload dollar disclosure + second guide raise toward $3.0B — evidence the AI tailwind is P&L, not narrative.

---

Catalysts & Risks

---

References

• MongoDB Q1 FY27 results — Atlas +29%, first GAAP profit, FY27 guide raised (8-K, 2026-05-28) · Q1 FY27 10-Q (SEC) — SBC $137.8M, RPO detail
• Quiver — MDB slides without fresh catalyst (2026-07-10) · Yahoo — Search + Vector Search GA for EA/Community, Voyage reranking preview (2026-07-02) · Investing.com — Needham PT to $430 (2026-07-10)
• Stack Overflow Developer Survey (Dec 2025) — PostgreSQL 55.6% usage vs MongoDB ~26%, only major DB with negative growth · CNBC — Supabase $10.5B, 60% of new DBs AI-created (2026-06-04)