MongoDB

MongoDB enters 2026 in the most uncomfortable spot a once-hypergrowth software company can occupy: still growing 20%-plus, still profitable, still gaining workloads — but no longer narrative-driving the AI database conversation it should be at the center of. CEO Dev Ittycheria spent 2024 and 2025 making a series of strategic bets that look correct in slow motion: the $220 million Voyage AI acquisition closed in February 2025 (best-in-class embedding and reranking models), Atlas Vector Search reached general availability, and Search Nodes were repositioned as the AI-native query layer.

The execution problem is twofold. First, Atlas growth decelerated from 38% (FY2024) to 32% (FY2025) to roughly 26% guided for FY2026, and the company has to convince the market this is the natural curve, not the beginning of a structural slowdown like Snowflake faced in 2023-2024. Second, the CFO transition — long-time finance chief Michael Gordon retired in mid-2024, with Serge Tanjga moving from interim to permanent CFO in 2025 — coincided with two billings misses that the market unforgivingly punished. The stock dropped from over $500 to under $230 across 2024-2025, the worst peer-relative drawdown of any large-cap database name.

The 2026 thesis distills to one question: does Voyage AI plus Atlas Vector turn MongoDB from "document database that also does search" into the default AI application backend? If yes, the FY2027 reacceleration to 25%+ revenue growth happens and the stock re-rates to 12x EV/sales. If no, MongoDB becomes a 15-20% compounder priced like a mature on-prem database company at 6-7x sales.

Institutional positioning on MongoDB has visibly bifurcated after the 2024 drawdown. Lone Pine Capital, Coatue Management and Tiger Global all hold meaningful positions but trimmed at different points — Tiger most aggressively, Lone Pine least so. ARK Investment Management exited substantially in 2023 and has not rebuilt. On the bullish side, Whale Rock Capital added meaningfully in late 2024 specifically citing Atlas Vector positioning, and TimesSquare Capital re-initiated after the Voyage AI deal closed. Quantitative shops including Renaissance and DE Shaw have rotated around the position multiple times — short-term volatility traders, not signal. The most informative non-presence: no major activist has emerged despite an unloved multiple, which is itself a vote of confidence in Ittycheria's strategic playbook. Hedge fund short interest fell from 7% in mid-2024 to roughly 4% entering 2026, indicating the bearish thesis is consensus rather than fresh.

MongoDB trades at approximately 7x forward EV/sales and 35x forward FCF entering 2026, a meaningful discount to Snowflake (11x sales), Datadog (12x sales) and a premium to traditional database peers like Oracle (4x) and IBM. The market is essentially pricing MongoDB as a mid-growth profitable software business rather than as an AI-database pure-play. Bull case (Voyage AI converts to 4-5pp Atlas reacceleration, FY2028 revenue $3.5B, FCF margin to 22%): fair value $400-460. Base case (FY2028 revenue $3.1B, growth holds at 22-24%, FCF 20%): $280-330 — close to current price. Bear case (Atlas decelerates to 18% by FY2027, hyperscaler vector pressure, FCF stalls at 18%): $180-210. The asymmetry has actually become attractive: meaningful upside in the bull case, modest downside in the bear case.

1. March 6, 2026: Q4 FY2026 / full-year FY2026 earnings (fiscal year ends January). Watch Atlas growth rate exit, FY2027 guidance shape (key question: 22% or 25%+ for FY2027 Atlas), Voyage AI workload attach commentary.
2. May 2026 (MongoDB.local World): Annual customer conference. Historical pattern: product launches that influence next 12 months of analyst models. Expect explicit AI agent-stack announcements, Atlas Vector v2 features, and likely a Voyage AI integration roadmap with metric goals.
3. September 2026: Q2 FY2027 earnings — first clean comparison quarter post-Voyage AI integration deliverables. Bull-thesis make-or-break: does AI workload mix on Atlas reach 15%+ of net-new bookings? Sell-side will recalibrate the FY2028 model on this data point.

MongoDB is the trade where I am most genuinely uncertain about the next 18 months but most genuinely confident about the next 5 years. The product strategy — owned embedding, vector + document in a single managed service, multi-cloud — is what every AI application backend should look like. The execution risk is real: Atlas growth deceleration is a hard math problem the company has to solve in plain sight, and the CFO transition created two quarters of damage that have not fully healed.

For position sizing, I would build into MongoDB through the FY2026 earnings cycle rather than load up before March 6. The post-Voyage AI workload data point in the September 2026 print is where the bull case is decided. If I see Atlas net-new attach to AI workloads above 15% by then, this is one of the higher conviction multi-year compounders in software. If not, the position thesis is wrong and I would exit clean rather than average down.