MongoDB's Q2 2026 Earnings Call: Contradictions Emerge on Atlas Growth, AI Adoption Timelines, and Infrastructure Positioning

The above is the analysis of the conflicting points in this earnings call

Date of Call: None provided

Financials Results

• Revenue: $591.0M, up 24% YOY and above the high end of guidance
• EPS: $1.00 per diluted share, up from $0.70 in the prior year
• Gross Margin: 74%, down from 75% in the prior year
• Operating Margin: 15%, up from 11% in the prior year

Guidance:

- FY26 revenue $2.34B–$2.36B (raised $70M); non-GAAP op income $321M–$331M; op margin high end 14% (from 12.5%); EPS $3.64–$3.73 on 87.4M shares; 20% tax.- Atlas H2 growth implied mid-20s%; non-Atlas subscription down mid-single digits; multiyear headwind $40M (from ~$50M); non-Atlas ARR to grow YoY.- Q3 revenue $587M–$592M; non-GAAP op income $66M–$70M; EPS $0.76–$0.79 on 87.7M shares.- Q3 op margin below Q2 due to lower non-Atlas and opex timing; expect low-20% YoY decline in non-Atlas.

Business Commentary:

Revenue Growth and Product Mix:* - MongoDB reported revenue of $591 million for Q2, up 24% year over year, exceeding the high end of their guidance. - The growth was driven by Atlas revenue growing 29% year over year, representing 74% of total revenue.

• Customer Additions and Atlas Performance:
• MongoDB added over 5,000 customers over the last two quarters, ending Q2 with over 59,900 customers.

• Atlas customer count grew to over 58,300, reflecting both new customers and existing customers deploying workloads on Atlas for the first time.

• Operating Margin Expansion:

• MongoDB achieved a non-GAAP operating margin of 15%, up from 11% in the previous year.

• The improvement in operating margin was attributed to revenue outperformance and a focus on disciplined investment.

• AI and Platform Adoption:

• MongoDB is emerging as a standard for AI applications, with a growing number of AI native startups and enterprises adopting their platform.
• The integration of search, vector search, and embedding models is attracting customers, particularly in the enterprise segment, where they are increasingly building custom AI solutions.

  Sentiment Analysis:

  • “We generated revenue of $591,000,000, up 24% year over year and above the high end of our guidance.” “Atlas revenue grew 29% year over year, representing 74% of total revenue.” “We are increasing our full year guidance across the board.” “We are raising our expectations for operating margin to 14% at the high end.”

  Q&A:

  • Question from Sanjit Singh (Morgan Stanley): What drove the strong sequential Atlas dollar adds and acceleration in Q2?
  • Response: Larger upmarket workloads are growing faster and longer; increased adoption of search/vector search; and strong new customer additions aided growth.
  • Question from Sanjit Singh (Morgan Stanley): Update on go-to-market and salesforce effectiveness after changes?
  • Response: Strategy unchanged: focus direct sales on large enterprises while self-serve covers SMB; not abandoning self-serve, and both motions are working well.
  • Question from Raimo Lenschow (Barclays): What’s driving self-service acceleration despite moving upmarket?
  • Response: Data-driven experiments, targeted outreach to SQL developers, and guided education (e.g., office hours) under an upgraded marketing team are boosting self-serve.
  • Question from Raimo Lenschow (Barclays): How should we think about EA/non-Atlas cohort dynamics into next year?
  • Response: Too early to call; outcome depends on Q3 FY26 multiyear dynamics, given last year’s strong multiyear compare.
  • Question from Tyler Radke (Citi): How much did AI use cases contribute to Atlas strength?
  • Response: AI-native customers are growing but were not a material driver; Q2 strength came from core customers and workloads.
  • Question from Tyler Radke (Citi): Is migration/app modernization velocity improving with new tools?
  • Response: Investing in AI-assisted tooling and delivery for app modernization; meaningful long-term opportunity, limited impact near term.
  • Question from Jason Ader (William Blair): Impact of Databricks Lakehouse/Lakebase and DocumentDB/Linux Foundation on MongoDB?
  • Response: OLTP is the AI strategic high ground; MongoDB’s JSON architecture plus integrated search/vector search is differentiated; hyperscaler clones are de-emphasized; partnerships remain strong.
  • Question from Jason Ader (William Blair): Why do many AI startups start on Postgres?
  • Response: Founders default to what they know; at scale, JSONB and performance limits emerge, prompting moves to MongoDB; MongoDB is investing in developer education/community.
  • Question from Mike Cikos (Needham): What explains Q2 Atlas consumption strength and large-customer momentum?
  • Response: Consumption grew 29% YoY with broad-based strength, notably among large U.S. customers; workloads are expanding longer; GTM changes helped; Q1 was a softer compare.
  • Question from Mike Cikos (Needham): Did multiyear outperformance reflect pull-forwards?
  • Response: No pull-forwards; strength was broad-based with more multiyear than expected; multiyear headwind reduced to $40M from ~$50M.
  • Question from Alex Sukin (Wolfe Research): Where is AI workload momentum and when will it materially impact growth?
  • Response: Architecture (JSON + search/vector + embeddings) resonates; enterprise AI is early; AI cohort not material yet, but positions MongoDB well for future demand.
  • Question from Alex Sukin (Wolfe Research): How are you balancing growth investments and margins?
  • Response: Revenue growth is the main margin driver; disciplined spending and reallocation sustain both growth and expanding margins.
  • Question from Kash Rangan (Goldman Sachs): Reconciling startup evangelism with enterprise-led growth; is DevRel a leading indicator?
  • Response: Current growth is led by large-enterprise workloads from moving upmarket; self-serve is effective; startup wins signal future but aren’t yet material.
  • Question from Brad Reback (Stifel): Outlook for EA growth given 7% ARR growth?
  • Response: Customers are embracing hybrid (on-prem + cloud); MongoDB’s portability offers flexibility; EA skewed to existing customers with optionality across environments.
  • Question from Ittai Kidron (Oppenheimer): GTM to capture large AI workloads; beyond self-serve?
  • Response: Scale self-serve until accounts warrant high-touch sales, then transition; enterprises remain early with low-stakes AI; cadence mirrors prior cloud motion.
  • Question from Ittai Kidron (Oppenheimer): Any change in multiyear vs annual mix and reasons?
  • Response: Mix not disclosed; multiyear usage is broader, not larger; rationale unchanged—strategy alignment and price lock-in for data gravity.
  • Question from Siti Panigrahi (Mizuho): When does AI adoption meaningfully contribute, given ROI concerns?
  • Response: Adoption will be gradual as output quality, security, reliability, and scalability improve; agents will increase system intensity—no single tipping point.
  • Question from Brad Sills (Bank of America): Where are R&D investments focused?
  • Response: Core platform performance (8.0/8.1) and expanding capabilities like vector/streaming; more details at Investor Day.
  • Question from Brad Sills (Bank of America): How much of the outperformance reflects targeting higher-quality workloads?
  • Response: A lot—upmarket focus is yielding larger, longer-growing workloads driving Atlas acceleration.
  • Question from Rishi Jaluria (RBC): MongoDB’s role in a multi-agent, natural-language future?
  • Response: MongoDB’s JSON model, integrated search/vector, and support for memory/orchestration position it well to store state, plan, and act in agentic systems.