Enterprise Spending Is Splitting in Two: Where the Money Went in Q2 2026

Two earnings calls this week told different stories. One reveals where enterprise budgets are going. The other reveals where they are not.

On July 14, IBM reported Q2 revenue of $17.2 billion — below the $17.86 billion analysts expected. Its Infrastructure segment fell 7 percent year over year. Consulting revenue was flat. CEO Arvind Krishna blamed clients shifting capital expenditure toward "supply constrained infrastructure" and said "numerous large deals failed to close on the timelines we expected."

Eight days later, Alphabet reported Google Cloud revenue of $24.8 billion — up 82 percent year over year. Its cloud backlog hit $514 billion. Alphabet raised its full year 2026 capital expenditure guidance to a range of $195 billion to $205 billion, up from $180 billion to $190 billion. In Q2 alone it spent $44.9 billion on AI infrastructure.

Two public companies. Seven days apart. Opposite directions. If you sell into enterprise accounts, this is not a news story. It is a prospecting map.

The money is moving from on premise to cloud, and the speed is accelerating

IBM's miss is not about a weak economy. The ISM Manufacturing PMI registered 53.3 percent in June, the sixth consecutive month of expansion. Corporate balance sheets are healthy. The issue is where CFOs are directing spend.

Traditional IT infrastructure — servers, storage, on premise software, consulting for legacy systems — is being squeezed. Enterprises are not buying less technology. They are buying different technology. Every dollar that shifts from an on premise line item to a cloud AI line item creates a winner and a loser. Google Cloud is the winner in this quarter's data. IBM is the loser.

For a sales rep, the practical question is not which stock to buy. It is which accounts to call. Every company that runs IBM's software or relies on traditional infrastructure consulting is a candidate for a migration event. When a legacy vendor reports a demand drop, their customers are not standing still. They are evaluating alternatives.

The $514 billion cloud backlog is a forward revenue indicator worth prospecting against

Google Cloud's $514 billion backlog represents signed contracts not yet delivered. That is multi year committed spend already on the books. Every company in that backlog has already decided to increase its cloud and AI investment. They have signed. They are waiting for delivery.

That delivery cycle creates procurement events. A company scaling its cloud infrastructure needs security audits, compliance reviews, data migration partners, architecture consulting, and monitoring tools. The cloud provider books the headline number. The companies that support that infrastructure capture the tail.

The same logic applies to every hyperscaler. Microsoft Azure and AWS report next week. If their cloud numbers show similar momentum, the opportunity set expands further. The signal is not which cloud provider is winning. It is that enterprise cloud AI spending is accelerating broadly, and every acceleration creates a vendor supply chain worth prospecting.

How to find the accounts that are actually spending

Three public data sources will surface the accounts in motion right now.

SEC filings from the hyperscalers. When Google, Microsoft, or Amazon raise capex guidance, the beneficiaries are not just the cloud providers themselves. They are the data center builders, the networking equipment vendors, the power infrastructure companies, and the professional services firms that support them. Pull the 10 Q and 10 K filings of the companies in that supply chain and look for revenue guidance increases, headcount additions, and facility expansions. Those are the accounts with budget and mandate.

Job posting data. A company that is scaling cloud infrastructure hires for specific roles: cloud architects, security engineers, compliance managers, and procurement specialists. Those job postings are public. An account that adds three cloud security job listings in a week is not thinking about buying. It is actively buying.

WARN notices and facility permits. Data center expansions require construction. Construction requires permits. Permits are public record. When a company files a permit for a new data center or a WARN notice for a facility expansion, it creates a 6 to 12 month window of related procurement. Every vendor that supports that facility — cooling, security, networking, power — has an entry point.

What this means for your pipeline this quarter

The Q2 2026 earnings season is creating a clear before and after. Companies that sell into traditional IT infrastructure will face a harder close environment. Companies that sell into cloud native, AI infrastructure, and the supply chain around hyperscalers will find buyers who have budget, mandate, and urgency.

If your prospect list is heavy on legacy IT accounts, it is time to expand upstream. Look at the customers of the hyperscalers. Look at the supply chain that supports them. Look at the job postings and the facility permits. The information is public. The buying signals are visible. The only missing piece is the rep who connects them.

SalesInt tracks these signals across thousands of companies so you do not have to monitor them yourself. Our weekly intelligence briefs surface which accounts are showing real buying signals based on public data. No guesses. No intent scores. Just filings, permits, job postings, and contracts that tell you who is ready to buy.