For three years, “enterprise AI adoption” was a phrase that appeared in every vendor presentation but rarely in financial results. Alphabet’s Q2 2026 earnings, reported on July 22, changed that.

Google Cloud hit $24.8 billion in revenue for the quarter — an 82% increase year-over-year. Cloud operating income reached $8.8 billion, up from $2.8 billion in the same period a year earlier. The cloud backlog now stands at $514 billion. These are not pilot numbers. They are the financials of a platform that enterprise customers are running real workloads on, at real scale.

Alphabet overall posted $119.8 billion in Q2 revenue, up 24% year-over-year and above analyst forecasts of $116.93 billion.

The AI Adoption Numbers Behind the Revenue

The revenue spike has a clear driver. Nearly 90% of the Fortune 100 are now using Gemini Enterprise. Gemini models are processing 22 billion API tokens per minute across Google’s infrastructure. The Gemini app has 950 million monthly active users.

These are not sign-ups or trials. Token consumption at 22 billion per minute means code is being generated, documents are being analysed, and agents are being run at industrial scale inside major organisations.

Cloud operating margins also improved dramatically, from roughly 10% a year ago to 35% this quarter. That shift tells a specific story: the unit economics of AI cloud are improving as demand grows, not deteriorating as sceptics predicted.

What Changed in 12 Months

Twelve months ago, most enterprise AI adoption looked like this: a handful of approved tools, a few internal pilots, a cautious procurement process. The conversations were about readiness, governance, and whether the ROI would actually show up.

Those conversations have not gone away — but the default answer to “should we adopt?” has flipped. Enterprises are no longer asking whether to use AI; they are asking which workloads to run where.

Google Cloud’s $514 billion backlog is the clearest evidence of this shift. Backlog represents committed future spend from customers who have already signed contracts. That number going up by more than $50 billion in a single quarter means the pipeline of committed enterprise AI spend is growing faster than the revenue is being recognised.

The Broader Picture

Alphabet’s results do not exist in isolation. Microsoft’s AI Copilot and Azure AI numbers have been similarly strong. Amazon Web Services reported enterprise AI workloads as one of the key growth drivers in its most recent results. The pattern is consistent: every major cloud platform is seeing enterprise AI pull through as real revenue.

The implication for business leaders is simple. If 90% of the largest companies in the world are running AI tools on your competitors’ infrastructure, standing still is a decision.

What This Means for Business

Enterprise AI has a revenue track record now. The “wait and see” posture becomes harder to defend when competitors are locking in multi-year cloud commitments and the infrastructure providers are posting their best growth quarters in years.

The practical question for most businesses is not whether to adopt AI but where to build capability and where to buy it. Building internal AI competency — the ability to evaluate tools, design prompts, manage agents, and interpret outputs — is the thing that determines whether you extract value from AI or just pay the vendor bills.

That is the exact gap that data training addresses. Teams that understand what the AI is actually doing, and can work with it strategically, consistently outperform teams that treat it as a black box. The numbers in Alphabet’s earnings are the destination; building the skills to get there is the work.

---

Enterprise DNA trains data professionals and business teams in AI, analytics, and the skills needed to build and operate AI systems. Explore the learning platform or talk to us about team upskilling.