John DiFucci

McGowan said colocation costs come ahead of revenue but accounting straight-lines both; ramp to full profitability is roughly a quarter, up to 5-6 months on the long end.

Ed said supply chain is secured with all goods expected within twelve months, majority in 2026. Contracts include mechanisms for price changes. Revenue model is set capacity with no usage component, straight committed deal over seven years. Revenue starts as capacity ramps, with partial year in 2026 and 2027, then full run rate for remaining six-plus years. Effectively looks like a subscription despite being consumption-related technology.

Tom said OpenAI has never been an Akamai customer. The partnership with OpenAI is limited to security vulnerabilities, helping define them and protecting customers from associated attacks. No impact on delivery or compute forecast.

Ed clarified the social networking consolidation deal is a different customer from the $200M AI deal. The $200M deal will be fairly even revenue, starting Q4 2026 as GPUs are deployed and customer testing begins. The prior $100M deal from last year started generating some revenue in Q4 2025 and will ramp throughout 2026. Some seasonal variation in compute tied to sports seasons but no major lumpiness.

Ed said weaker cash flow was a timing issue from cash receipts, payments, and large year-end tax payments. For the $55M restructuring, a little over half was intangible asset impairments with no cash impact. Severance (less than half) will be paid out in Q1 2026.

Tom said transformation will be largely done by beginning of next year. Continued hiring through at least first half of next year for increased hunting capacity in security and compute.

Tom confirmed four major competitors gone since pre-pandemic. Better pricing environment, improved traffic trends. Looking at mid-single-digit decline going forward with long-term goal of stable delivery revenue.

Tom listed new channel partners: Presidio, WWT, Optiv, GuidePoint, DefenseX, IBM, Deloitte, Macnica, OCD, Accenture, Carso. Ed added 80-90% of new logo acquisition comes through channel, over half of security business is channel-driven.

Tom confirmed new customer sales often lead with Guardicore or API security but the whole security platform can then be included and grown. Partners are effective at cross-selling the full portfolio.

Ed broke down the path: delivery drag moderating, security maintaining 10% growth from newer products, compute reaccelerating to 20% CAGR led by cloud infrastructure.

Tom cited three advantages: efficient data movement/lower egress, distributed platform for performance, and managed containers in 700+ cities. Ed added four synergies: backbone, operations, engineering, and go-to-market.

Hillary framed operating margin as the North Star and said gross margin should flatten as the ramp finishes, but deferred specific guidance to the October investor day.

Hillary gave a full reconfirmation of the long-term CAGR targets, citing the building RPO as support.

Clayton attributed the CapEx increase to timing, not component prices, and described pass-through cost mechanisms that protect Oracle's margins when prices rise.

Mike confirmed halo effect across apps (AI embedded), OCI budget creation for customers, and Sovereign AI pipeline. Doug deferred FY27 CapEx specifics but highlighted decoupling of CapEx from Oracle capital requirements.

Mike argued Oracle's complete suite strategy (back-office, industry, front-office) with baked-in AI differentiates against best-of-breed competitors. Deferred revenue up 14% vs 11% in-quarter growth supports acceleration thesis.

Larry emphasized inferencing demand as much larger than training, with unprecedented calls for any available capacity worldwide. Also highlighted non-AI growth from companies like TIMU.

Safra clarified Stargate is still forming and existing OpenAI business is relatively small. Larry added that if Stargate materializes as advertised, RPO growth is understated.

Safra explained hyperscaler partners are highly motivated to deploy quickly but timing is not fully in Oracle's control. Larry said primary/backup coverage in key regions within ~12 months.

Safra confirmed OCI gross margins continue to improve but did not provide a specific updated number. Attributed improvement to software-driven automation and operational leverage as a core software company.

Aneel said Workday is arriving now via hard, deeply embedded agents; Gerrit and Gabe framed it as a continuous, long-term loop of adoption and ROI.

Garrett cited the Travel Agent and Sana for ITSM as examples of AI enabling Workday to expand beyond traditional HR and finance. Aneel added that AI resets competitive boundaries and Workday can make bets in new markets leveraging its robust data model.

Zane explained Q1 sequential decline from DIA contract not continuing, fewer days in Q1, and some Q4 deal pushes. Gross retention remains at 97%. Guidance philosophy unchanged. Pushed back on analyst's suggestion of additional conservatism.

David confirmed 150 deals and 22+ partners since October launch, with a strong pipeline of additional partners looking to build on GPC.

Carl confirmed AI selling (30%+ attach), full suite (30%+ new wins, 50%+ in focus industries), and partner sourcing (15%+ of ACV) are all paying off. Record partner impact and create-and-close momentum.

Rene emphasized that AI cannot replace physical chips and that enterprise AI adoption is still in very early stages. Cited Google's $180 billion CapEx as evidence of unprecedented demand. Called AI the 'final frontier' in productivity transformation and expressed strong long-term excitement.

Rene argued lower inference costs expand compute demand into power-constrained endpoints where Arm dominates. Noted no hyperscaler is pulling back on AI spend.

Rene guaranteed v9 mix will be considerably higher than 25% a year out, driven by PCs, the flagship-to-midrange waterfall, and all-v9 data center; the flat quarter was a mid-market growth mix effect.

George said it is still early innings, deepest among developers, with security adoption front-run by AI adoption and not yet enterprise-wide.

Burt explained the temporary ARR-revenue divergence: CCPs with extended time hit ARR immediately but revenue impact lags, creating a short-term decoupling.

Nikesh detailed a measured integration: protect the top line, take out overlap for margin, then deliver new product capability.

Nikesh drew cloud security analogy where numbers lagged 2-3 years. Prisma AIRS following XSIAM trajectory but volume not yet there. Urged patience given $600 billion data center build-out still underway.