Your MTTR is a lagging indicator of your change process
Most outages we trace back aren't monitoring failures. They're change failures that monitoring caught late. Fix the handoff, and the dashboard improves on its own.
Research · white papers Ten years of the market, read from the operator's chair.
// Written for client briefings. Every claim cites a public source; where we give our own assessment, we say so.
Security New · Sep 2026
Security Meets Observability, 2016–2026
How the security analytics market turned over, who bought whom, and why the SOC and the NOC now run on one data layer. Includes an agentic SOC simulation and autonomy tiers for security actions.
$32B Google’s Wiz acquisition, closed March 2026: cloud security joins the data platforms Request a copy → Observability Updated · Sep 2026Modern Observability & DEX, 2016–2026
The 2026 platform landscape, hyperscaler tools, Nexthink and DEX, RUM and synthetics, ThousandEyes and its rivals, AppDynamics then and now, and a minute-by-minute agentic incident.
~$3.0B Palo Alto’s Chronosphere purchase: security vendors now buy observability Request a copy → Event ManagementEvent Management, End to End
How a raw signal becomes a routed, resolved and learned-from incident, and where observability and ITSM have to be designed as one system.
1 system observability and ITSM delivered together, not as two projects Request a copy → SecOpsModern SecOps in the Age of AI
Building or upgrading a SOC where AI agents handle triage under governance and analysts keep control of high-impact decisions.
Minutes from first access to lateral movement, which is why triage has to run at machine speed Request a copy →Field Notes
Original thinking, not recycled vendor talking points.
// Written from the operator's chair and the channel's side of the table, not from vendor press releases.
The cheapest telemetry is the telemetry you decided not to collect
Ingest-based pricing punishes indecision. A telemetry budget is a design decision, not a finance problem to fix at renewal.
Why the new office always goes live without monitoring
Procurement closes the PO, the network team closes the ticket, and nobody owns day two. Monitoring belongs on the bill of materials.
What your reseller knows about your renewal that you don't
Deal registration, quarter-end timing and distribution programs shape your price more than your usage does. Here's how to read the board.
The CMDB is the AI project
Agents act on topology and ownership data. A stale CMDB doesn't slow an agent down; it makes it confidently wrong, faster. Fix service models and ownership before you buy autonomy.
Autonomy is earned one action class at a time
Let an agent restart a pod long before it can roll back a release. Grant each right with approvals and an audit trail, and take it back automatically the first time it fails.
Measure the tickets that were never opened
Faster ticket closure is a service-desk metric. Problems fixed on the device before anyone calls are the metric executives feel, and that is what DEX and self-healing move.
Security is becoming a use of observability data
The SOC and the NOC now collect the same logs, flows and traces. Collect once, normalize once, and route each team what it needs. Paying twice to ingest the same event is a design flaw, not a budget line.