When data grows faster than budgets, customers start sampling

Observability costs rise with the amount of data a company produces: logs, metrics and traces from every system it runs. When that data grows faster than the budget for monitoring it, engineering teams respond by keeping less of it, and the vendor's revenue slows with them.

Sampling is a budget decision

Teams can filter logs, shorten retention or sample traces rather than keep everything. Each of those choices cuts cost without dropping the vendor, so it never shows up as churn. It shows up as slower growth from customers who are otherwise satisfied.

Usage-based pricing makes this visible faster than a seat-based model would.

What I'm watching

• Net revenue retention, which captures reduced usage
• Commentary on customer cost optimisation
• Growth in newer products that customers adopt alongside monitoring

Where I stand

Short, on data volumes outrunning budgets and teams responding by keeping less. The product remains essential; the bill is where the tension shows.

What would change my mind

Net retention rising while customers are still cost-conscious. That would mean new products are growing faster than sampling cuts the old ones.