Agentic AI triples telemetry data, pushing observability costs past compute budgets

Telemetry data growth outpaces enterprise readiness

Omdia research highlights a rapid surge in telemetry data volumes driven by agentic AI workloads. Enterprises reported an average 3.7x increase in telemetry data over the past 12 months, with nearly half anticipating growth between 6x and 100x within the next two years. On average, organizations expect a 9.5x increase in data volumes from agentic AI alone.

The research, cited by TechRadar, indicates that 69% of enterprises running agentic AI projects now face observability costs that exceed their compute and infrastructure budgets. For context, the average annual observability spend has reached $3.17 million, growing at 28% year-over-year. Nearly one in five enterprises already spends over $5 million annually on observability, with 35% seeing budget increases exceeding 52%.

Costs force delays in AI deployments

The financial strain is already affecting AI adoption. Fifty-nine percent of enterprises have terminated or delayed at least one agentic AI deployment due to unsustainable monitoring costs. Projects paused or canceled include critical functions like cybersecurity monitoring, compliance, and fraud detection—areas where 27% and 26% of respondents, respectively, reported disruptions.

Keir Walker, a Senior Market Research Analyst at Omdia, said the scalability of telemetry pipelines is the core issue, with no signs of data growth slowing. Torsten Volk, a Principal Analyst at Omdia, described the current phase as "the beginning of the hockey stick" for telemetry data growth, emphasizing that enterprises are still underprepared for the surge.

Telemetry pipelines as a solution

Enterprises with mature telemetry pipelines are 50% more likely to be prepared for data growth and 80% more likely to avoid operational cost challenges. The research found that 70% of organizations are actively evaluating telemetry pipeline solutions, with nearly one in four prioritizing performance over existing vendor relationships.

Only 35% of organizations report widespread agentic AI deployment today, signaling that adoption is still in its early stages. However, the data suggests that early investment in scalable telemetry infrastructure could mitigate the cost and operational risks associated with AI-driven data growth.

What this means

Lazyfounder analysis — our interpretation, not reported fact.

This research underscores a critical tension for enterprises investing in agentic AI: the technology’s promise is being undermined by the unforeseen cost of monitoring its outputs. For founders and operators, the takeaway is not to scale back AI ambitions but to treat observability as a first-order concern—not an afterthought. The fact that 59% of enterprises have already delayed or canceled AI projects due to monitoring costs should serve as a warning: without a strategy to manage telemetry data at scale, even well-funded AI initiatives can become financially unsustainable.

The data also signals an opportunity for startups building telemetry pipeline solutions. The 70% of enterprises now evaluating such tools are prioritizing performance over vendor loyalty, suggesting a market ripe for disruption. For operators in DevOps or cloud infrastructure, the numbers make a clear case for auditing observability spend early—before it eclipses compute costs and forces tough trade-offs in cybersecurity and compliance.

Related on Lazyfounder

• More technology news
• AI Hype Meets Reality: Debunking Claims, Privacy Risks, and Geopolitical Rivalry