Big Tech earnings preview: Amazon, Alphabet and Microsoft face the $1 trillion AI question

Microsoft: can AI growth justify the spending?

Microsoft sells business software and operates Azure, one of the world's largest cloud platforms.

Azure revenue grew 43% in the June quarter. For the upcoming quarter, management expects growth of 44% to 45% in constant currency, meaning exchange-rate movements are excluded.

Demand appears strong, but growth comes at a cost. Microsoft expects quarterly capital expenditure, or spending on long-term assets such as data centres, to exceed 50 billion USD.

The question is whether growing Azure demand and paid adoption of tools such as Microsoft 365 Copilot can offset the rising costs of delivering AI services.

Alphabet: can cloud profits catch up with growth?

Alphabet owns Google, whose businesses include Search, YouTube and Google Cloud.

Google Cloud revenue surged 82% in the June quarter, making it the fastest-growing of the three cloud businesses. Investors will now watch whether operating profits can keep pace as the company adds expensive computing capacity.

Alphabet also faces a second challenge. AI-powered search could encourage more searches and new advertising opportunities, but it could also change how users interact with traditional advertisements.

For Alphabet, the test is therefore twofold: turn cloud growth into stronger profits while protecting the advertising business that helps finance its AI ambitions.

Amazon: where is the cash?

Amazon combines its global retail platform with Amazon Web Services (AWS), the cloud business that provides much of its operating profit.

AWS revenue grew 37% in the June quarter, its fastest expansion in several years. Yet Amazon's free cash flow, the money remaining after operating expenses and investments, was negative 7.6 billion USD over the preceding 12 months.

This does not necessarily signal a struggling business. Amazon generated substantial operating cash, but infrastructure investments absorbed even more.

For investors, the distinction matters. Growing revenue is encouraging. Growing profits are better. Eventually, those profits also need to translate into cash.

The trillion-dollar question is about returns, not revenue

AI infrastructure is expensive to build, but potentially profitable to operate once customers start using it.

The difficulty is timing. Companies spend billions on chips, buildings and electricity connections today, while the revenue they hope to generate may arrive over several years.

This creates an important distinction between earnings and cash generation.

A company can report excellent profits while spending more cash than it produces. Equally, negative free cash flow during an investment phase does not automatically indicate poor long-term returns.

For shareholders, the key is whether future profits eventually justify today's spending.

Valuations add another layer. When share prices already reflect ambitious growth expectations, companies have less room for disappointment.

That means even impressive quarterly results might trigger a negative market reaction if management raises spending forecasts without offering convincing evidence of future returns.

Risks: when growth becomes expensive

Three risks deserve particular attention.

First, AI demand could grow more slowly than the infrastructure built to serve it, leaving companies with underused capacity.

Second, competition could push down prices for computing services, limiting profitability even as customer adoption increases.

Third, rising electricity, financing and equipment costs could prolong the period before investments generate meaningful cash returns.

The early warning signs include slowing cloud growth, declining margins and repeated increases in spending forecasts.

Investor playbook: look beyond the earnings headline

• Compare growth with profitability. Faster cloud revenue growth is more valuable when operating margins remain healthy.
• Follow the cash. Track whether free cash flow improves as earlier investments begin generating revenue.
• Watch spending guidance. Rising investment is easier to justify when expected demand and returns rise alongside it.
• Check portfolio concentration. Several technology holdings may depend on the same AI investment cycle.

The bill is arriving. Will the profits follow?

For much of the AI boom, investors have focused on how quickly companies could build bigger, faster and more powerful computing systems. That remains important, but the next chapter looks increasingly financial.

Microsoft must show that growing AI adoption supports its expanding infrastructure. Alphabet needs to translate exceptional cloud growth into sustainable profits while protecting Search. Amazon must demonstrate that substantial investments can eventually generate more cash.

None of these questions will be settled by one earnings report. And spending heavily today is not necessarily a mistake if the eventual returns justify it.

The AI opportunity is becoming easier to see. The cost of pursuing it is also becoming harder to ignore. Ultimately, shareholders do not earn returns from how much companies spend, but from what that spending produces.