Q2 Tech Earnings Preview: It's All About AI ROI

Big tech is on deck as earnings season ramps up. With many Magnificent Seven companies living up to their new "lag seven" moniker in recent months, the stakes are high—even though revenue and earnings per share (EPS) growth estimates have cooled.

The biggest questions facing this group are all too familiar, but this earnings season could bring some of the clearest answers yet. First, investors will want to see Alphabet (GOOGL), Microsoft (MSFT), Amazon (AMZN), and Meta Platforms (META) provide evidence that their AI spending sprees are paying off and their cloud businesses can continue growing rapidly. Apple's (AAPL) hardware and services units will also be under the microscope as usual, with investors focused on the impact of high memory prices on margins.

Robotaxi progress, automotive margins, and any robotics (Optimus) updates will be in focus when Tesla (TSLA) reports July 22 amid talk of a potential SpaceX (SPCX) merger. Investors will have to wait until late August, however, to see results from AI bellwether Nvidia (NVDA), leaving chip demand dynamics somewhat opaque for now.

Importantly, the Mag Seven's core businesses haven't shown any signs of faltering in recent quarters even though cash flows have been strained by massive AI spending. That trend will need to continue with the group's combined capital expenditures (CapEx) set to exceed $700 billion in 2026. Ultimately, for big tech investors this quarter, the main question remains: Will the record spending pay off?

"I don't think anyone is arguing that AI won't be revolutionary for the global economy, but the economics tied to heavy spending on developing LLMs are still up for debate, especially given the shift by enterprises from 'tokenmaxxing' toward 'token efficiency' over the past couple months," said Nathan Peterson, director of derivatives research and strategy at the Schwab Center for Financial Research. "Hyperscaler CapEx guidance will be key this quarter, along with evidence of a return on investment from that spending."

Seemingly ever-increasing AI spending has raised the bar for the Mag Seven this quarter, with investors now looking for clearer evidence of meaningful returns.

Combined, the Mag Seven's CapEx is expected to surge toward $1.1 trillion in 2027, according to estimates from Goldman Sachs (GS) and Morgan Stanley (MS). The record spending spree was easier for investors to stomach when it was funded purely by high-margin core businesses producing abundant free cash flows (FCF), but now it's begun to pressure cash flows and push some companies to the debt markets.

The Mag Seven's combined debt sales have surpassed $130 billion so far this year, and FCF for the group has declined sharply since its 2024 peak. Take Amazon, for example. Its trailing 12-month FCF declined 95% year over year to just $1.2 billion in the first quarter. Aggressive AI spending has also weighed on FCF at Meta, Alphabet, and Microsoft. If that trend continues this quarter, there will be significant pressure on the Mag Seven to prove their AI investments are at least beginning to generate a strong return on investment (ROI).

Metrics that could provide insight into the Mag Seven's AI ROI are likely to be intensely examined. Copilot adoption and Azure revenue growth attributable to AI will be in the spotlight at Microsoft, while Gemini App and Gemini Enterprise growth will be key at Google.

"I suspect the growth and/or adoption rates around these AI business segments will be scrutinized similar to cloud growth forecasts and the trajectory of that growth," Peterson explained, noting that investors will be hoping to see revenue and user growth accelerating.

Amazon Web Services' AI revenue run rate, Meta's AI advertising tool adoption, and Tesla's full self-driving sales and subscriptions will also provide a read on the Mag Seven's AI ROI. And any updates on current or upcoming AI-linked returns in big tech conference calls will be closely parsed. The stakes of the largest spending spree in history continue to rise with Mag Seven cash flows under pressure and debt burdens growing.

After a period of exceptional expansion, revenue and EPS growth for many of the Mag Seven are expected to moderate this quarter. With cash flows under pressure and debt burdens rising, the key question is whether management teams will remain confident enough in future AI demand to maintain record levels of capital spending amid the growth slowdown.

Overall, the info tech sector is expected to report an impressive year-over-year earnings growth rate of 63.3% in the second quarter, according to FactSet data. That's well above the S&P 500's estimated year-over-year earnings growth rate of 23.6%. However, this rapid earnings growth rate is being driven mainly by the semiconductor and semiconductor equipment industry, which includes high-flying memory makers like Micron Technology (MU) and SanDisk (SNDK). Analysts are forecasting far more pedestrian earnings growth from the Mag Seven.

Consensus sees year-over-year EPS growth of roughly 16% for Microsoft, 8% for Amazon, and 1% for Meta. Alphabet, Tesla, and Apple are expected to post EPS growth in the low-to-mid 20% range, however, leaving them roughly in line with the broader market. Still, the EPS trend for the Mag Seven is a far cry from the epic growth rates seen in recent years.

With earnings growth expected to be weaker, investors may start to question whether the Mag Seven can—or should—continue their record spending.

"It's been a common theme for several quarters, and investors have mostly exhibited patience around the heavy spending, but the hyperscalers have lagged the market in 2026, so investment discernment has been growing," Peterson said. "It comes down to AI spending guidance versus AI ROI evidence. If there is clear, tangible evidence that the heavy spending is paying off, investors likely won't care about the level of spending, at least for now."

The performance of Mag Seven companies' core cash-flow engines in the upcoming quarter will be critical as they assess the prospects for ongoing AI spending. For many of these companies, that means the cloud business. Big tech cloud businesses provide high-margin, recurring revenue streams and are the main way these firms convert enterprise and consumer AI spending into near-term profits.

So far, there's little evidence of a slowdown here. In the first quarter, Google Cloud and Microsoft Cloud posted 63% and 29% year-over-year revenue growth, respectively, while Amazon Web Services reported 28% year-over-year sales growth. The performance of these businesses has been so strong that it has attracted a new competitor, Meta.

Bloomberg reported July 1 that Meta is developing plans to launch a new cloud infrastructure business, called Meta Compute, in an attempt to monetize its internal infrastructure. The business unit will sell excess AI computing capacity, positioning it as a direct competitor to Amazon Web Services, Microsoft Azure, and Google Cloud.

Investors will want to closely track the growth rates in these cloud businesses this quarter, with a focus on forward guidance. Any signs of a slowdown ahead could weigh on the group's ability to keep spending on AI and potentially impact share prices.

Whether AI spending continues will also depend on the overall demand for computing power. Early July earnings from AI bellwether Taiwan Semiconductor Manufacturing (TSM) seem to indicate AI compute demand remains strong. The company posted a 68% year-over-year jump in revenue in the second quarter and noted production capacity for its advanced 3nm (nanometer) chip manufacturing process is fully booked through 2028. Still, compute demand has been insatiable throughout the AI boom, and many investors are beginning to wonder how long that trend will continue.

"Because insatiable demand for compute has essentially become the expectation, the risk could be skewed to the downside if one of the hyperscalers lowers CapEx guidance," Peterson said, noting that if Mag Seven companies cut their spending plans this quarter, it "might draw a broader question of the Large Language Model investment thesis."

As companies race to deploy AI, attention is increasingly shifting from making the best AI models and services toward a new challenge: making the technology affordable enough for widespread adoption.

Rising token costs have emerged as a major issue for consumers and businesses alike this year. Although major AI developers have been pushing for efficiency, multiple business leaders have commented on the high costs of implementing AI. Palo Alto Networks CEO Nikesh Arora, for example, argued in a July 9 interview with CNBC that more efficiency gains are necessary if enterprises are going to start "betting the farm on AI evolution."

"We need to see the pricing for AI come down," Arora said. "The pricing for AI today makes it very hard for enterprises to take the bet."

Arora's comments, which are emblematic of the outlook of many CEOs, raise a serious concern for Mag Seven companies. While AI demand remains strong, this group will now have to deliver AI services and cloud hosting at the right price. And if enterprises are reluctant to pay up for AI, margins and sector valuations could be at risk. This means investors may want to look for evidence of Mag Seven companies pushing for AI efficiency or reducing what they charge for AI products and related services this quarter.

Many AI companies, including some of the Mag Seven, have already begun to respond to pricing issues. On the consumer side, Alphabet slashed the price of its entry-level AI subscription from $7.99 to $4.99. OpenAI, meanwhile, is reportedly considering significant price cuts for its AI services as it seeks to compete with rival Anthropic, The Wall Street Journal recently reported. OpenAI and other AI developers are also looking to make their models more token-efficient. CEO Sam Altman recently highlighted that his company's latest model is 54% more token-efficient for agentic coding, but CEOs are seemingly thirsty for more efficiency gains.

Meanwhile, Microsoft is seeking to replace OpenAI and Anthropic models with in-house AI models to lower costs, leaning into the "model independence" trend. Meta CEO Mark Zuckerberg also recently cited his company's model independence as a strength, arguing Meta will be able to offer some AI services at lower prices.

Even as AI startups and the Mag Seven focus on efficiency, AI costs are rising as usage grows, leading some businesses to route workloads through aggregator platforms that allow the use of cheaper models like China's DeepSeek.

Rising AI competition and a more cost-conscious mindset from businesses threaten the near-term profitability of AI startups but could also squeeze Mag Seven cloud and AI-service margins moving forward. For investors, this makes guidance and earnings call updates on these businesses particularly important this quarter.

While gauging the ROI of AI investments, tracking CapEx guidance, and watching cash flows and cloud margins will be critical for tech investors this earnings season, there are a few other company-specific factors worth monitoring. Consider the following:

• Software sales growth amid potential AI disruption to software as a service (SaaS) business models
• Gaming hardware and content sales growth during the division's struggles with cost reductions, game studio spin-offs, and layoffs
• PC/hardware revenue growth and margins with high memory and component costs acting as a drag

• Advertising business growth rate and margins as well as updates on the use of new AI advertising tools like the Seller Assistant
• Implementation and benefits from the fulfillment efficiency and robotics push
• Updates on e-commerce margins, the impact of tariffs, and consumer spending trends

• Core search and advertising revenue growth amid the changing AI search landscape
• YouTube paid subscriptions and ad revenue growth
• Growth metrics and expansion initiatives in the Waymo robotaxi business

• Core advertising business and impact of AI-based tools for advertisers like the Meta Advantage+ Creative suite
• Updates on the launch of the new AI cloud business, Meta Compute, and the manufacturing of the new data center chip, Iris
• Profitability in the augmented and virtual reality business, Reality Labs, which lost more than $4 billion in the first quarter alone
• Any discussion of the adoption of Meta's AI coding agent, Muse Spark, which is meant to compete with Anthropic's Claude Code

• iPhone sales momentum to justify the upgrade "supercycle" narrative, particularly demand in China
• Monetization strategy for the new generative-AI-powered Siri amid an expected paid subscription tier and regulatory roadblocks in Europe and China
• Updates and guidance for hardware margins amid surging memory and component prices

The Mag Seven have largely thrived during the AI buildout, with revenues surging amid record demand for their products and services. However, with EPS growth rates slowing—and stubborn inflation, a hawkish Federal Reserve, and geopolitical tensions acting as a headwind—beating Wall Street's forecasts and offering strong guidance this quarter may be as important as ever. For major tech firms reporting this week, analysts* expect the following:

GOOGL: Reporting July 22 after the close. Earnings per share (EPS) is expected to rise 24.9% from a year ago to $2.88, and revenue is expected to jump 21.1% year over year to $116.7 billion.

TSLA: Reporting July 22 after the close. EPS is expected to rise 25.5% from a year ago to $0.5, and revenue is expected to jump 13.9% to $25.6 billion.

MSFT: Reporting July 29 after the close. EPS is expected to rise 16.3% from a year ago to $4.24, and revenue is expected to jump 14.6% year over year to $87.6 billion.

META: Reporting July 29 after the close. EPS is expected to rise 1.1% from a year ago to $7.22, and revenue is expected to jump 26.6% to $60.2 billion.

AAPL: Reporting July 30 after the close. EPS is expected to rise 20.5% to $1.89, and revenue is expected to jump 15.5% to $108.6 billion.

AMZN: Reporting July 30 after the close. EPS is expected to rise 8.3% to $1.82, and revenue is expected to jump 17.1% to $196.4 billion.

*Note: Average analyst earnings and revenue estimates are as of July 16, 2026, courtesy of Schwab.com, and are subject to change.