Advantage+ and advertising efficiency: Meta’s AI-powered advertising suite was reported to have reached an annualised run rate of approximately US$60 billion. Markets may focus on whether efficiency gains are being sustained and whether the platform continues to attract new advertiser spending.

Reality Labs costs: Virtual and augmented reality operations generated operating losses estimated to have exceeded US$16 billion in earlier periods. Markets look for spending discipline or signs that costs are beginning to moderate.

Compute-leasing potential: Reports indicate Meta is discussing leasing data centre capacity to external AI developers, with one potential arrangement reportedly valued at up to US$10 billion.

Capex escalation versus free cash flow: Full-year 2026 capex guidance of US$125 billion to US$145 billion may test investor confidence if revenue growth does not clearly outpace component cost inflation.

AWS re-acceleration: AWS revenue grew 28% YoY in Q1 2026. With a leadership transition underway, markets focus on whether the business is re-accelerating towards the 30% range as enterprise AI workloads migrate to the cloud.

Custom silicon commitments: Graviton, Trainium and Nitro processor platforms reached an annualised run rate of ~US$20 billion. Trainium 3 capacity is reported nearly fully subscribed, helping reduce dependence on third-party silicon margins.

Retail margin resilience under tariff pressure: Amazon’s value-led pricing protects retail volumes. Prime Day activity across 23 countries during Q2 provides a near-term indicator of global consumer spending conditions amidst tariff impacts.

Project Kuiper and capex trajectory: Full-year capex is tracking towards ~US$200 billion. Q2 results absorb roughly US$11 billion in sequential satellite production and launch outlays.

Edge AI hardware and the iPhone upgrade cycle: Apple Intelligence uses a hybrid processing model that completes AI tasks locally. Shipment data suggested iPhone 17 unit growth in China of ~24% YoY, testing whether that momentum translates into reported revenue.

Greater China market share: China’s approval of Alibaba’s Qwen model locally reduced regulatory friction. Markets watch if Apple sustains market-share gains against domestic competitors like Huawei.

Global memory costs and gross margin: Advanced-memory tightness pressures component costs for premium inputs. Management guided gross margin of 47.5% to 48.5% for the quarter.

Foldable iPhone production and Services growth: Supply chain data points to increasing production targets for a foldable iPhone model towards 10 million units, while Services approaches a US$30 billion quarterly revenue run rate.

Google Cloud backlog conversion: Google Cloud’s contracted backlog exceeded US$460 billion in the prior quarter, while revenue grew 63% year on year (YoY). The focus is on how quickly that backlog converts into recognised revenue.

Search and YouTube advertising resilience: The bulk of Alphabet’s revenue remains tied to advertising. More than 30% of customers’ Search spending now uses AI-enabled campaigns such as Performance Max.

Capex and margin trajectory: Alphabet’s higher infrastructure investment may place pressure on free cash flow as the company expands its AI capacity. Markets want evidence that this spending produces measurable revenue.

US Department of Justice antitrust proceedings: The DOJ has secured remedies addressing search distribution and distribution agreements. Management commentary on implementation and potential additional relief may influence market pricing parameters.

Azure growth compared with capex: Microsoft expects Azure revenue growth of 39% to 40% in constant currency for the quarter, while investing approximately US$190 billion in capex during calendar year 2026. The focus is on whether this spending is translating into cloud revenue.

Agentic AI and Copilot adoption: Following Microsoft Build 2026 and the introduction of Microsoft Scout, traders will scan the earnings call for evidence of commercial traction. Paid seat growth and ARPU will indicate whether enterprise adoption is expanding.

Gross margin direction: Higher infrastructure costs associated with data centre expansion may place pressure on software-driven margins. Whether gross margin expands or contracts will be an important metric.

Regulatory backdrop: The FTC has examined partnerships between cloud providers and AI developers, including Microsoft and OpenAI. Management commentary on regulatory developments may affect longer-term growth assumptions.

Blackwell and Vera Rubin supply constraints: Demand for Nvidia’s Blackwell architecture and the announced Vera Rubin platform remains a key focus. An important variable is advanced packaging capacity, including TSMC’s CoWoS process capacity crunch, which could limit the pace of the production ramp, even if underlying demand remains strong.

Gross margin sustainability: Nvidia delivered a generally accepted accounting principles (GAAP) gross margin of 74.9% in the prior quarter. Analysts will monitor whether pricing, product mix and production costs can sustain margins near this level as advanced architectures enter larger-scale production and hyperscalers expand their own custom silicon programs.

Sovereign AI demand: Governments across Europe, the Middle East and Asia-Pacific are investing in domestic AI infrastructure. These customers may have different procurement priorities, project structures and implementation timelines from major hyperscalers. Growing sovereign AI orders could broaden Nvidia’s addressable market and reduce its reliance on a small group of large customers.

Custom silicon competition: Microsoft’s Maia platform, Alphabet’s Ironwood tensor processing unit, Amazon’s Trainium and Meta’s custom accelerators could reduce hyperscaler dependence on Nvidia hardware over time. The pace at which these alternatives affect graphics processing unit procurement remains a long-term variable. Management commentary on data centre demand, backlog and forward commitments may provide further evidence.