데브 이체리아의 최고경영자(CEO) 임시 복귀와 충격적인 수장 교체 속에서도 몽고디비, 아틀라스 성장 목표치 20대 중반으로 상향
뉴욕 나스닥 마켓사이트 인베스터 데이 — 2026년 연 9월 29일
몽고디비(MongoDB)는 연례 인베스터 데이를 개최하여 경영진의 갑작스러운 퇴임이라는 악재와 이례적으로 공격적인 장기 재무 목표 상향이라는 두 가지 상반된 소식을 동시에 소화했습니다. CJ 데사이(CJ Desai) CEO가 메타(Meta)로의 이전을 위해 지난 금요일 전격 사임함에 따라, 이사회는 후임 정식 CEO를 물색하는 동안 창업자이자 전임 CEO인 데브 이체리아(Dev Ittycheria)를 임시 사장 겸 CEO로 선임했습니다. 경영진은 이번 수장 교체가 행사 전체의 빛을 바래게 하는 대신, 향후 3년간의 아틀라스(Atlas) 매출 성장률 목표를 기존 20%대에서 20대 중반(mid-20%)으로 끌어올리는 도약대로 삼았습니다. 이와 함께 자체 관리형 제품인 엔터프라이즈 어드밴스드(Enterprise Advanced, EA) 사업의 두 자리 수 성장 공약, 그리고 아틀라스 에이전트 엔진(Atlas Agent Engine) 및 아틀라스 인피니트(Atlas Infinite)를 비롯한 대규모 AI 인프라 신제품군을 새롭게 발표했습니다.
전격적인 사임, 그리고 다시 지휘봉을 잡은 창업자
마이클 베리(Michael Berry) 최고재무책임자(CFO)는 행사의 서두에서 이번 경영진 교체 이슈를 정면으로 다루며, 회사는 "8-K 공시에 담긴 내용 외에 추가로 언급할 수 없다"고 선을 긋는 한편 "몽고디비의 사업은 평소와 다름없이 흔들림 없이 진행되고 있다"고 강조했습니다. 데사이가 재임한 약 9개월 동안 이사회 자리를 지켰던 이체리아 임시 CEO는 전임자의 성과와 한계에 대해 가감 없이 털어놓았습니다. 이체리아는 "CJ가 잘한 점은 회사의 영업력을 최고 경영진(C-level) 수준으로 확실히 끌어올린 것"이라며, 데사이가 서비스나우(ServiceNow) 시절 구축한 인맥을 활용해 주요 핵심 고객사와의 계약 체결을 가속화했다고 평가했습니다. 다만 이번 성과가 "원맨쇼는 아니었다"고 선을 그었습니다. 그는 1년 만에 두 번째 CEO 교체를 겪으면서 조직 내 혼란이 가중될 것이라는 우려에 대해, 지난 수년간 회사가 "수많은 임원진 변화를 겪어왔으며" 현 경제 환경에서는 "변화가 곧 상수"라고 애널리스트들에게 일축했습니다.
Raised Long-Term Targets Are the Real Headline for the Model
Buried beneath the CEO news was a materially more constructive long-term financial framework. MongoDB shortened its guidance horizon from five years to three and raised its Atlas growth target to the mid-20% range, up from the 20%-plus target set a year ago. Total revenue growth is now targeted at 20%-plus (versus "high teens" previously), and for the first time management put a specific floor on EA and Other growth at 10%-plus, up from an implied mid-single-digit assumption last year. Berry was explicit that these are not ceilings: "100 to 200 basis points was not a ceiling" on operating margin, and if enterprise AI adoption, AI-native contribution, or the new agentic products outperform, "hopefully, we can grow Atlas faster." The company also disclosed it hit every single commitment made at last year's Investor Day, with fiscal 2026 Atlas growth of 29%, EA growth of 7% (guided to 11% for fiscal 2027), operating margin of 18.5% (guided to approximately 21%), and free cash flow conversion of 108%. On a Rule of 40 basis, the company guided to Rule of 44 this year and set an implicit ambition of approaching Rule of 50 by the end of the three-year window, though Berry declined to formally commit to that figure.
Atlas Agent Engine: A New Monetization Layer Built on Existing Contracts
The most significant product announcement was Atlas Agent Engine, described by AI products co-chief Pablo Stern-Plaza as "one of the largest expansions of our platform in company history." The offering bundles persistent agent memory, built-in security guardrails, and model/cloud-agnostic infrastructure into what MongoDB is positioning as the "agentic run time" layer sitting on top of its existing database. Crucially, it requires no new sales motion: "All existing Atlas customers are already entitled to it. No new contract, no new sales motion, all Atlas builders can start using it today," Stern-Plaza said, meaning tens of thousands of existing Atlas customers already have access to the public preview. A live customer testimonial from Citizens Bank underscored the product's early traction in regulated industries. Vikas Agarwal, the bank's SVP and Head of Data Platforms, said the memory layer resonated because "when you go and directly hit the memory layer, it's pretty faster... quite faster than hitting to the actual LLMs," and praised the platform-agnostic design that avoids lock-in to any single model provider — an increasingly important consideration as banks toggle between Claude, Codex, and other LLMs on a near-monthly basis. Partner Erica Volini added that one systems integrator saw "984% higher query speed and a 63% latency drop when testing Agent Engine against sovereign AI offerings" in internal deployments. Management does not expect meaningful revenue contribution until fiscal 2028.
Atlas Infinite Separates Compute From Storage for Bursty Agent Workloads
Chief Product Officer Ben Cefalo called Atlas Infinite "Atlas' biggest architectural innovation since its inception in 2016." The product decouples compute and storage, allowing customers to scale each independently — a direct response to what Cefalo described as the fundamentally different traffic pattern of agentic workloads compared to user-driven applications: "Agents don't generate steady, predictable traffic. They can sit idle for long stretches and then suddenly introduce massive unpredictable demand that hits the database all at once." Early preview customers including Icon Solutions and PicPay reported handling multiples of their normal peak traffic. The product just entered public preview on AWS only; broader cloud availability and GA are not expected until early fiscal 2028, with Berry cautioning "don't look for meaningful revenue in '28, but certainly in '29."
MongoDB 9.0 Delivers the Largest Performance Jump Yet, and Jevons Paradox Is the Bet
The freshly released MongoDB 9.0 delivers up to 20% higher transactional throughput, 35% faster reads, and 30% faster updates versus version 8.0, which itself is running on nearly 80% of the Atlas fleet just two years after launch. Management's underlying thesis is explicitly Jevons paradox: faster, cheaper performance doesn't cannibalize consumption, it expands it. "When we introduce better price performance, we've seen stronger growth in subsequent periods," Cefalo said, pointing to the fact that 8.0's rollout coincided with accelerating, not decelerating, Atlas revenue growth. Ittycheria reinforced the point directly in the Q&A, noting that when the market worried 8.0 would let customers "use less gas," instead "they wanted to drive that car more because they put more workloads on it because they saw so much benefit."
EA Re-Acceleration Is Real, Not Just Vector Search Noise
Enterprise Advanced, MongoDB's self-managed offering, has quietly become a second growth engine. After averaging mid-single-digit growth for years, EA ARR growth has now been double-digit for three consecutive quarters — notably, before the June general availability of Search and Vector Search for self-managed environments, which only just began contributing. Berry highlighted that roughly two-thirds of EA customers with more than $100,000 in ARR also use Atlas, directly rebutting the bear case that EA growth comes at Atlas's expense: "we see that cohort actually growing faster than the total... more Mongo is better across either EA or Atlas."
AI-Native Revenue Still Small, But the Growth Curve Is the Point
Management disclosed for the first time that AI-native and Frontier Lab customers currently represent just 3% to 5% of Atlas revenue, but that cohort's revenue is compounding at 165% year-over-year, with customer counts having doubled over two years. Berry was unusually blunt about managing expectations here: "As we sit here today, AI natives make up less than 5% of Atlas revenue. I'm just going to cut the question off. When you think of that, think of 3% to 5%." The message to investors is that this is an emerging, high-growth but still immaterial layer of the business today — one of several explicit upside levers to the raised Atlas target, alongside enterprise AI adoption, application modernization, and new product monetization.
Application Modernization Partnership With Cognition Signals Renewed Push
After what Ittycheria acknowledged was "a bit of a quiet period" on the application modernization platform (AMP) over the past year, MongoDB announced a new partnership with AI coding company Cognition to accelerate migrations of legacy Oracle and SQL workloads. Ittycheria admitted the company had underestimated how long AI model maturity would take to solve the hardest part of migrations — rewriting brittle application logic and stored procedures — but said recent model progress has changed the calculus: migrations that once took years can now happen "in days and weeks." CRO Ryan Ban quantified the opportunity at over $10 billion in incremental TAM unlocked by AI-driven modernization tooling, calling it one of the most underappreciated growth vectors in the model.
Capital Return: Another $1 Billion Buyback
Berry announced the board approved an additional $1 billion buyback authorization, funded by free cash flow that reached almost $500 million last year on 108% conversion. The company holds over $2 billion in cash and reiterated it will pursue only "disciplined" tuck-in acquisitions for technology and talent, similar to the Voyage AI deal, rather than large-scale M&A. Management also committed to GAAP profitability starting in fiscal 2027, a threshold it had not previously guided to explicitly.