Eric Sheridan

Jassy gave specific adoption metrics on perishables, Whole Foods and Daily Shop, though he did not break out differences by geography.

Jassy confirmed continued significant capital investment, framing AI as a once-in-a-lifetime opportunity. Cited 28% AWS growth on $150B run rate and unique chip positioning with Trainium and Graviton.

Jassy confirmed backlog $244 billion up 40% year over year and 22% quarter over quarter with lot of deals in pipeline. Vast majority of capital and capacity consumed by external customers with Amazon always a large but small fraction of total. Internally over a thousand AI applications deployed or building from Rufus to Alexa Plus to fulfillment forecasting to advertising optimization. Added 3.99 gigawatts of power in last twelve months with 1.2 gigawatts in Q4 alone. Expects to double capacity again by 2027 and continue growing in the ballpark of current rate.

Jassy said over 1 million robots in fulfillment network with expectation of more over time. Robotics improve safety boost productivity increase speed and let human teammates focus on problem solving. Expects future fulfillment network where robots and humans complement each other. Will continue to advance and invest very significantly in robotics helping on safety productivity speed and cost.

Jassy said priorities are broadest selection at lowest prices with fast delivery. Done forward inventory buys. 2 million-plus sellers with diversity meaning not all will pass tariffs. Diversified production away from China over past 6 years. Olsavsky noted Q2 includes SBC step-up and Kuiper costs with wider guidance range for uncertainty.

Jassy expressed impression with DeepSeek training techniques and inference optimizations. Said everyone building frontier models is working on same things with a lot of leapfrogging. Drew parallel to early cloud days where lower per-unit costs lead to more total spending. Believes inference cost decline will be very positive for customers and AWS business.

Ariane Gorin split AI into two buckets: in-product AI already lifting conversion (recommendations, ranking, personalization) and newer natural-language experiences that deepen traveler intent understanding for future conversion. She framed AI experiences like ChatGPT, Claude and Google as new paid-channel opportunities against the two-thirds direct booking base.

Ariane detailed clear positioning for each brand: Expedia as one-stop shop, Vrbo as trusted VR brand, Hotels.com relaunched. Marketing mix balances upper funnel brand spend with improving performance spend returns.

Ariane confirmed positioning work largely complete across all three brands. Cited international growth opportunity (room nights growing faster outside U.S.) and continued marketing effectiveness improvements.

Ariane outlined growth from new partners plus existing partner expansion, cited geographic diversity (65% outside U.S.), multiple partner types, and emerging LOBs like car rentals and advertising.

Ariane highlighted portfolio diversification, AI traffic opportunities, and B2B's geographic diversity. Scott noted strength in direct traffic and growing app conversion.

Ariane outlined four AI pillars: enhancing products (review summaries, property highlights), driving traffic through AI search, B2B partnerships with AI-native companies, and team effectiveness.

Strauss distinguished mobile (light, ~9-minute sessions) from console (immersive) as consumption types rather than devices, expecting technology to broaden the market for both.

Strauss described daily optimization of UA models with tight payback period guidelines. Noted post-Apple attribution challenges have been surmounted as evidenced by Zynga results. Said new AI-driven efficiency opportunities exist.

Strauss confirmed DTC share has been growing but declined to disclose percentage or targets. Emphasized balancing DTC margin benefits with third-party retailer marketing value, positioning it as all of the above.

Karl praised Netflix as a terrific partner, mentioned BioShock movie in development with them, but emphasized media deals are nice-to-have not must-have.

Strauss noted the Big 4 sports are competitive and current portfolio offers sufficient growth opportunity without expansion.

Strauss emphasized the existential importance of new IP, noting hit ratios are high across both mobile and console portfolios.

Mertz declined a specific 2027 guide but pointed to a track record of investing behind growth while still expanding margins, implying a floor on reinvestment capacity.

Brian distinguished API hosts (need better tools) from primary home hosts (need frictionless listing). AI enables automated listing creation from photos and address.

Ellie explained Brazil scaled in a handful of years with continued incremental market share gains. Japan started just a year ago with local marketing for domestic stays. Each market has different timeline but consistent approach yields results.

Brian said marketing will bundle homes, services and experiences under single brand. Shifting from TV to social media. Ellie clarified $200M is for field operations and supply acquisition, not programmatic.

Brian said AI engineering productivity gains not yet a step change but expects 30% increase medium-term. Ellie confirmed capital allocation unchanged: invest, M&A where relevant, return capital.

Jamie outlined continued investment in focus categories, AI, eBay Live and vehicles. Highlighted AI-driven marketing efficiencies including low-cost creative generation and 40% more CRM engagement from AI-written subject lines.

Jamie confirmed lower funnel marketing efficiencies from competitive dynamics and outlined a multipronged agentic strategy: doubling down on unique non-new inventory, value-added services like authentication, and building proprietary AI capabilities including an AI shopping assistant.

Jamie framed eBay as relatively resilient given 40%+ used/refurbished inventory, noting historical pattern of consumers gravitating to value and sellers being nimble adapting to changes.

Jamie highlighted collectibles innovation with PSA/TCGplayer, P&A expansion with vehicles via Caramel, and fashion as the next major focus category opportunity.

Jamie noted more affluent consumers remain strong (luxury, collectibles) while less affluent consumers are more pressured in this environment.

Baszucki framed competition broadly as the need for play and creation, expecting AI to let everyone build games (2D to 3D) and redefine the category.

Baszucki said age-check lets Roblox target 18+ players (roughly 80% of the global gaming market, monetizing ~1.5x), and higher DevEx incentivizes novel games.

Baszucki said transparent, long-term-oriented discovery has broadened the healthy diversity of the content mix, with Moments as a future complement.

Baszucki said older cohorts are drawn in by viral new-genre hits (Dress To Impress, Grow a Garden), and the tech roadmap aligns to sports/racing/RPG genres.

Baszucki described a monetization menu (freemium, virtual currency, paid access, ads, physical shopping); Chopra framed it as a user-growth flywheel.

Ewout highlighted Level 2/3 mix reaching high 50% of room nights, correlated with higher direct booking rates and frequency; Glenn signaled additional Genius enhancements for 2026.

Glenn emphasized the goal is comprehensive cross-vertical supply rather than any single gap, while Ewout highlighted payments as strategic underpinning and the economic benefit of single acquisition costs.

Ewout cited lower-than-average but positive ROI channels being scaled, declined to share competitive details, and noted the mix effect would moderate by Q2.

Ewout confirmed no material shifts in investment strategy; continuing with transformation program, GenAI, fintech, and connected trip buildout regardless of near-term macro changes.

Steve detailed predictive AI/ML for consumer-carrier matching and generative AI for product and agent tooling while keeping headcount lean; direct.

Steve said it is both, but market forces driving carriers into the soft-market growth cycle matter most, expected to last 2 to 3 more years, with MediaAlpha's AI and platform tools amplifying it.

Pat emphasized the lean 146-person cost base and the 2022-2023 downturn playbook of belt-tightening while keeping the core team intact.

Management pointed to Medicare, agent-based carrier partnerships, and data science investment as key growth and efficiency initiatives.

Evan pointed to 85% other-revenue growth and subscription headroom below 3% penetration as diversification; direct.

Evan gave a broad execution and agentic-AI vision but few specific rollout milestones; partial.

Evan described developer traction from the 2024 dev version and the breadth of experiences planned at launch; direct.

Evan said Snap invests where it can differentiate (on-device image/video and 3D generation) and leverages its developer ecosystem; direct.

Ravi confirmed $100M investment on track with expected costs. Both executives highlighted velocity and quality benefits from unified tech stack.

Tony said DashPass is a single-digit % of potential with current couple-times-a-week vs 20-25 eating occasions. Adding non-restaurant benefits, in-store dining, and Going Out as new DashPass value drivers.

Tony emphasized product improvement across many small vectors: catalog depth, accuracy, affordability, Dasher matching. Customers graduating from top-up to larger baskets.

Two-part AI question; von Ahn cited record content output and more fluid video call, Munson tied gross margin toward ~69% as AI use in the product expands; direct.

Luis said a game-like visual app drives engagement a text chatbot cannot, but they will adapt if consumer behavior shifts to aggregators. Direct.

Skaruppa said the key margin cost is the marginal cost of Max (a good sign it is growing), with faster hiring and marketing secondary; AI costs could improve faster.

Wilson described bundle as successful last year with even more uptake expected. Long-term vision for deeper connection between products. International growth potential cited with NFL fandom growing globally, referencing Australian interest.

Canfield said OpEx is flat except for Battlefield go-to-market costs. R&D incremental is slight as studios near end of cycle. Post-launch investment also expected. Discipline maintained in overall cost allocation.

Wilson contrasted with prior Battlefields where community alignment was insufficient. This time engaged hundreds of thousands in map construction, weapon lineup, and progression feedback. Expanding to wider audience and Asia.

Peters noted positive retention impact from gaming but acknowledged small acquisition effect to date; investing in IP-based games, TV games, and kids experiences.

Sarandos sees AI as better tools for great artists, not replacement; Interpositive acquisition accelerates proprietary filmmaking GenAI. Peters added AI improving recommendations and ads.

Sarandos emphasized repeat business from creators as the ultimate success indicator; highlighted landing competitive projects and managing customer-competitor dynamics with studios.

Management cited high contribution margins and continued operating leverage; direct.

Management quantified Vector's record January and 10% sequential Q1 expectation; direct.

Management expects cloud cost as a percentage of revenue to fall over time; direct.

Alex (on for Eric Sheridan) asked about Hatch investments and structural margins; Schwarzbach said Yelp is investing across go-to-market and product/engineering to scale Hatch from a startup and expects its long-term margin profile to resemble other subscription businesses, with continued near-term investment to capture the market opportunity.

Schwarzbach said macro pressure is more significant on RR&O, with advertiser caution rising into Q4, while services advertiser demand stays consistent.

Nachman addressed RRNO strategy directly, framing services as the priority while keeping Yelp positioned to capture RRNO demand when the macro turns.

Adam confirmed global ex-China TikTok bid focused on operational control of algo and data, describing APP's AI model expertise as unique qualifier.

Adam said personalized generative AI ads are still R&D; tricky for full-screen video vs text ads, but expected to be impactful in 2025.

Jen said Reddit is media-agnostic on outcomes and is easing video porting via Memorable AI, but video is not a focus.

Steve pointed to machine translation and in-country community work as the international user runway, with revenue to follow.

Robins framed predictions as mostly incremental in non-OSB states (CA, TX, FL), likening it to opening the rest of the country to a strong sports product.

Gwynne cited $6B+ in Q2 government contracts, sticky enterprise revenue (never lost an enterprise customer), and large aviation/maritime headroom; Elon expects enterprise revenue to exceed consumer.