Third Quarter 2026 Earnings Call - DUOLINGO INC

Financial Performance:

• Q3 Daily Active Users (DAU) grew 29% year-over-year, accelerating from 23% in Q2.
• Subscription bookings and revenue growth accelerated, with bookings growing double-digits in most regions.
• Full-year 2024 guidance: 12% bookings growth, 18% revenue growth, and 26.5% adjusted EBITDA margin.
• Q4 2024 guidance: 13% bookings growth, 14% revenue growth, 26.1% adjusted EBITDA margin, and roughly flat adjusted EBITDA at $84 million.
• Gross margin expected to decline slightly in Q4 (approx. 1 point) due to the rollout of video call to all paid subscribers, though overall margins remain attractive.
• $11 million bonus accrual recorded in Q3 for re-accelerating user growth; additional portion included in Q4 guidance, with full payment expected in early 2027.
• Paid marketing spend is approximately $100 million (60% of non-comp marketing) and remains profitable with improving LTV/CAC ratios.

Guidance and Future Outlook:

• Management anticipates Q4 DAU growth to remain similar to Q3 levels.
• Full-year 2027 outlook will be provided in February.
• Strategy for 2027 involves maintaining user growth focus while shifting emphasis toward monetization without degrading the free experience.
• Ambition remains to reach one billion learners, with chess reaching 10 million DAUs.
• Long-term goal is to leverage AI to make the platform as effective as a one-on-one human tutor while maintaining a gamified experience to ensure retention.

Business Segments and Product Lines:

• Duolingo Max Sunset: Duolingo Max is being phased out; existing Max Individual users are migrated to Super Family at a similar price point, while Max Family is expected to end by year-end.
• Video Call Rollout: Video call (formerly Max-only) is being moved to Super Duolingo for all paid subscribers, enabled by reducing AI inference costs from ~$0.30 to under $0.01 per call via open-weight models.
• Chess: Reached 10 million DAUs; identified as a significant under-monetized segment with future monetization opportunities.
• Math: Growing at a pace similar to languages, with plans to expand content; currently less popular than chess but viewed as a massive opportunity.
• Light Tier: Currently in testing as a lower-cost ad-supported option; adoption has not been a "home run" yet, with potential regional rollout to lower GDP markets.
• Gaming Partnerships: Recent collaborations with Brawl Stars and Eggie Party drove user acquisition; partnerships are cross-promotional (no cash exchange) and expected to occur a few times annually.
• New Courses: AI integration has accelerated course development, reducing time from idea to launch (e.g., Chess in 9 months).

Market and Competitive Landscape:

• DAU growth accelerated across eight of the ten largest markets globally, indicating broad-based geographic expansion.
• Competition for consumer attention remains high (e.g., TikTok, Instagram); Duolingo's strategy relies on gamification and habit formation to retain users against these distractions.
• Management emphasizes having the largest dataset on language learning, providing a competitive moat for AI model training.
• Inference cost deflation (shifting to open-weight models) is reducing barriers to scaling AI features across the user base.

Risks and Challenges:

• AI Costs vs. Usage: Balancing the rollout of AI features (which increase AI usage) against cost deflation; while costs are dropping, total spend on free user inference is estimated in single-digit millions for next year.
• Monetization Friction: Guardrails against adding friction to the free user experience could limit traditional monetization levers, requiring innovation in A/B testing for conversion.
• Operational Scaling: As experiment velocity increases 90% year-over-year, bottlenecks are shifting to idea generation and bug resolution by non-engineers.
• Customer Migration: Risk of cannibalization when moving Max users to Super Family, though management reports this is not material to date.
• Market Saturation: Potential difficulty in maintaining high retention if the "one-on-one" AI efficacy target is not met relative to human tutors.

Management Commentary and Tone:

• Tone is highly confident and optimistic, characterized as "bullish" regarding the billion-learner goal and AI capabilities.
• Management expressed pride in executing the strategy to prioritize "teaching better" and user growth, validating the delayed monetization focus.
• Luis von Ahn (CEO) emphasized that the majority of growth stems from product improvements (the "green machine") and retention enhancements rather than just partnerships.
• Jillian Munson (CFO) highlighted the efficiency gains in marketing and AI cost management, noting that paid marketing is profitable and ROI is improving.
• Management explicitly stated a strict internal guardrail: no degradation of the free user experience moving forward.

Analyst Questions and Answers:

• DAU Drivers: Analysts asked about the split between partnerships and product improvements; Management confirmed the majority of growth is from product optimization and retention improvements, with partnerships providing targeted boosts.
• Subscription Growth Lag: Analysts noted slowing paid sub growth vs. DAU growth; Management explained this is a lag effect due to prior bookings mix and extended free trials, with sub growth expected to catch up as bookings rebuild.
• Max Sunsetting: Questions on the impact of sunsetting Max; Management confirmed it is a net positive, improving revenue and expanding AI access to 10x more users, with minimal short-term bookings impact.
• Gross Margins: Analysts inquired about margin guidance post-video call rollout; Management confirmed a slight sequential decline in Q4 but expects better-than-expected margins for the full year due to lower AI costs.
• Marketing Efficacy: Questions on LTV/CAC; Management reported strong performance in paid marketing, with spending decisions being data-driven and highly profitable.
• AI Experimentation: Analysts asked about the volume of A/B tests; Management revealed a 90% YoY increase in tests, enabled by AI allowing non-engineers to deploy changes, with plans to run multiples more tests in 2027.
• Long-term AI Role: Questions on whether AI will become core to the learning model; Management believes they can replicate human tutor efficacy within a few years but will maintain a gamified interface to prevent churn.

Other Key Points:

• The company is transitioning from a "user growth only" phase to a balanced approach prioritizing both growth and monetization in 2027.
• Direct payment integration in apps (outside of the web) is being expanded, particularly for performance marketing acquisitions.
• Ads business is transitioning to a higher-quality direct sales model, expected to drive growth in the medium term.
• Management indicated that "idea generation" is becoming the primary bottleneck as process efficiency improves, rather than engineering capacity.