Datadog launches Experiments

Datadog made Datadog Experiments generally available, combining A/B testing, product analytics and observability with AI‑oriented tools like a Bits AI Security Analyst and an MCP server for embedding live telemetry in AI workflows. Analysts have reacted with upgrades while some large investors trimmed positions, reflecting mixed market sentiment even as the company pushes observability into AI governance. (simplywall.st) (investing.com) (marketbeat.com)

Datadog has put a new product called Experiments into general release, folding A/B testing into the same system many customers already use to watch apps, logs and user behavior. (datadoghq.com) The company said on April 2 that Datadog Experiments is now available “to customers everywhere,” letting teams design, launch and measure tests inside Datadog instead of stitching together separate analytics and monitoring tools. Its product page says the service tracks behavioral data, application performance and business metrics in one workflow. (datadoghq.com 1) (datadoghq.com 2) A/B testing splits users into groups to compare two versions of a feature, and observability tools watch what the software does after a change goes live. Datadog’s documentation says Experiments can use Real User Monitoring, Product Analytics and warehouse data, and can tie into Datadog Feature Flags to run randomized rollouts. (docs.datadoghq.com) (datadoghq.com) Datadog has spent 2026 adding artificial intelligence tools around that same telemetry layer. On March 9, it launched a Model Context Protocol server that feeds live logs, metrics and traces into coding agents such as Cursor, Claude Code and OpenAI Codex while keeping Datadog’s access controls in place. (datadoghq.com) (docs.datadoghq.com) On March 23, Datadog said its Bits AI Security Analyst became generally available for Cloud Security Information and Event Management users, and said the tool can cut threat investigation time by more than 90%. The product page describes it as an always-on assistant that investigates alerts and sends results to tools such as Slack and Jira. (datadoghq.com 1) (datadoghq.com 2) The pitch is that one change to a product can now be judged on several fronts at once: whether users click more, whether revenue moves, and whether the software slows down or throws errors. Datadog’s blog says the goal is to let product and engineering teams run “trustworthy experiments” without waiting on separate data or reliability teams. (datadoghq.com) Wall Street’s reaction has been mixed. Yahoo Finance, citing Investing.com, reported on April 12 that Guggenheim upgraded Datadog to Buy, and separate coverage said the firm set a $175 price target after arguing the market was underestimating Datadog’s artificial-intelligence-related growth. (finance.yahoo.com 1) (finance.yahoo.com 2) Some large holders moved the other way. MarketBeat reported on April 13 that Baillie Gifford cut its Datadog position by 65.3% in its fourth-quarter filing, selling 7,451,970 shares and ending the period with 3,961,509 shares worth about $538.7 million. (marketbeat.com) Datadog is trying to make the system that watches software into the system that also decides whether new code worked. Experiments is the clearest version of that push so far: one dashboard for shipping a change, measuring the result and checking whether the app stayed healthy. (datadoghq.com) (datadoghq.com)

Get your own daily briefing

Scout delivers personalized news, insights, and conversations tailored to your role and industry.

Download on the App Store

Shared from Scout - Be the smartest in the room.