Data Layer Investments Key for AI Growth in Second Half of 2026, Analyst Says

Investment focus for the second half of 2026 should center on the data layer of the software stack, according to Heath Terry, head of AI investment research at Citigroup. Speaking on CNBC’s “Squawk on the Street,” Terry highlighted companies with consumption-driven business models as prime candidates for AI-related growth.

“The companies that are winners are the ones that are exposed to consumption,” Terry explained. “AI is driving massive growth in consumption, particularly at the data layer.”

This segment, positioned above storage and below applications, handles data cleaning, formatting, and organization for artificial intelligence use. Software companies delivering increased access to maneuverable data are becoming crucial components of the AI ecosystem.

Notable companies in this space include Snowflake, MongoDB, Datadog, and Elastic, all showing strong revenue growth. Snowflake recently reported 33% annual revenue growth with shares trading at 112 times earnings for the next 12 months and an enterprise value exceeding 14 times sales. Datadog posted 25% annual revenue growth with a forward price-to-earnings multiple of 98 and an EV-to-sales ratio of approximately 19.

The current “agentic” phase of AI development—characterized by multiple software agents coordinating tasks—has influenced hardware design toward enhanced processing unit coordination. At the application level, this translates to model routing systems that optimize efficiency by directing workloads to appropriate models.

Private companies like Not Diamond, Martian, and OpenRouter are developing routing solutions to help enterprises reduce AI implementation costs. This optimization becomes critical as companies grow more selective about AI computing usage, measured in token consumption.

Investor confidence in sustained data demand is driving a renewed focus on software sector differentiation after earlier widespread devaluation, informally termed “SaaS-mageddon.”

Source link