Should Datadog kill its Real User Monitoring module?

Direct Answer

No — but the steelmanned case for killing RUM is more credible than Datadog bulls want to admit. Real User Monitoring is almost certainly sub-5% of Datadog's ~$3.0B+ ARR run-rate, it loses the developer-pull war to Sentry, and it sits in a category Microsoft Clarity has commoditized to free for 80% of use cases. The kill case has four legs that all hold weight: weak revenue, weak developer mindshare, AI-agent-driven UIs compressing the human-UX-monitoring TAM, and a free competitor anchoring price expectations to zero. But killing RUM outright would cost Datadog more than the $120-150M of revenue it walks away from — it would break the APM+RUM+Synthetics platform story, hand Bits AI a missing context layer for end-user-impact correlation, and signal weakness in a quarter where the narrative is already "is Datadog still a platform or just APM with bolt-ons." The right move is a refocus, not a kill: shut down mobile RUM (where Sentry/Embrace win on developer love), double down on web RUM tied to APM traces, and reposition the module as AI-agent-impact monitoring — the one wedge where Datadog's distributed-tracing DNA actually beats the field.

The Steelmanned Case For Killing RUM

• Sub-5% revenue, low NRR contribution. RUM is bundled into the "observability" SKU bucket Datadog never breaks out, but channel checks and Forrester's 2025 Digital Experience Wave place it materially behind APM, Logs, and Infra. At ~$120-150M ARR against a ~$3.0B+ run-rate, the module is rounding-error revenue with rounding-error growth contribution — and it dilutes gross margin because real-user telemetry volume is brutal.
• Sentry dominates developer-first frontend monitoring. Sentry crossed $200M ARR with a developer-pull motion Datadog cannot replicate from the top-down. Frontend engineers install Sentry on day one of a project; they install Datadog RUM only if their platform team forces it. That's a structural loss, not a marketing problem.
• AI-agent-driven apps reduce human-UX-monitoring TAM. As 2026-2028 application surfaces shift toward agent-driven flows (Claude, GPT, Gemini agents calling APIs, headless browser sessions, voice-first UIs), the share of revenue-critical interactions that involve a human watching a page render compresses. RUM's core measurement — how a human experiences a page — becomes less of the funnel.
• Microsoft Clarity is free and good enough for 80% of use cases. Clarity ships heatmaps, session replay, and rage-click detection at $0, backed by Microsoft's distribution. For mid-market customers who want "what's broken on the checkout page," Clarity ends the conversation. That anchors the price ceiling for the entire category.

The 4 Reasons To Keep RUM

• Cross-sell glue for APM customers. Roughly 30-40% of APM customers also use RUM, and the joint-deal economics matter at renewal — bundled discounts on RUM lift APM seat counts and reduce churn risk. Killing RUM unbundles a story the sales team uses every week.
• Synthetic + RUM combined story holds enterprise. Enterprise buyers (Fortune 500 retail, banking, airlines) want one vendor for synthetic checks, real-user monitoring, and APM correlation. Splitting that story means losing the seven-figure platform deals to Dynatrace or New Relic, both of which still own the integrated narrative.
• Bits AI investigation needs RUM context for end-user-impact correlation. Datadog's AI-investigator only earns its keep if it can answer "which users were impacted, on which pages, with what business consequence." Without RUM, Bits AI investigations end at the trace boundary — useful for SREs, useless for the VP of Digital who actually signs the check.
• Killing it signals weakness; competitors reposition. The day Datadog announces RUM EOL, Dynatrace's CMO writes the blog post: "Datadog gives up on user experience." That narrative bleeds into APM and Logs renewal conversations regardless of merit. Strategic retreats in observability get punished, not rewarded.

What Killing RUM Would Cost

• $120-150M of ARR walked away from, plus the unbundling discount lift on APM at renewal — call it $180-220M of revenue impact in year one.
• Loss of APM-RUM-Synthetics platform narrative, which is the single biggest reason enterprise CIOs choose Datadog over a best-of-breed stack of Sentry + Pingdom + Honeycomb.
• Bits AI investigation becomes blind to end-user impact, weakening the AI-observability story exactly when AI-observability is the only growth narrative the market wants to hear.
• Talent and IP write-down — RUM employs 200-300 engineers, many with deep frontend-instrumentation expertise that doesn't transfer cleanly to other modules.
• Customer-trust signal that Datadog "kills modules," which makes every future module purchase a riskier decision for buyers who have to defend the choice internally.

The Refocus Alternative — Kill The Wrong Parts

• Kill mobile RUM. iOS and Android session monitoring is where Sentry, Embrace, and Instabug are eating Datadog's lunch on developer love. The mobile RUM business inside Datadog is sub-$30M ARR and structurally losing. EOL it, hand customers a Sentry migration path, take the small revenue hit.
• Double down on web RUM tied to APM traces. The unique wedge is end-to-end correlation: page-load slow → trace ID → backend service → root cause. That's a story Sentry cannot tell at the same fidelity. Invest the engineering capacity freed up by mobile RUM EOL into deepening trace-to-RUM correlation.
• Reposition RUM as AI-agent-impact monitoring. When an AI agent calls your API, what does the "user experience" look like? Latency budgets, error patterns, retry storms, hallucination-correction loops. Datadog's distributed-tracing DNA is the only RUM platform positioned to own this category in 2027.
• Kill the standalone RUM SKU. Bundle RUM into APM at zero incremental list price. Lose the line-item revenue, gain the platform stickiness, raise the floor on APM ASPs.
• Cut the session-replay arms race with Microsoft Clarity. Don't try to win on heatmaps and rage-click counts. Cede the marketing-analytics edge of RUM to Pendo and Clarity; keep the engineering-correlation edge.
• Re-org RUM under the APM GM, not as a standalone product line. Signals internally and externally that RUM is a feature of the platform, not a product fighting for its own roadmap.

What Datadog Should Actually Do

• Announce a RUM refocus on the next earnings call, not a kill. Frame it as "AI-era user experience" and tie the narrative to Bits AI.
• EOL mobile RUM with a 12-month migration window and a Sentry partnership announcement — turn the retreat into a partnership story.
• Ship RUM-for-AI-agents in H2 2026 as the flagship feature: agent session tracking, tool-call latency budgets, hallucination-correction loop detection.
• Tighten the APM-RUM-Synthetics bundle with a single "Digital Experience" SKU that simplifies the sales motion and raises the platform ASP.
• Acquire or partner for product-analytics overlap (Heap, June, or a smaller PostHog-style asset) to close the Pendo gap without building it organically.
• Re-org under APM GM and stop treating RUM as a peer product line.

Strategy Options

Strategic Option Flow

Bottom Line

The answer is no, do not kill RUM — but the bullish-platform narrative is masking a real strategic problem. RUM is a weak module in a commoditizing category, losing developer-pull to Sentry and price anchoring to Microsoft Clarity. The right play is a surgical refocus: kill mobile RUM, bundle web RUM into APM at zero list, and reposition the module as AI-agent-impact monitoring before someone else owns that category in 2027. Do that, and RUM stops being a defensive line item and becomes the wedge into the next observability platform cycle. Do nothing, and in three years RUM is a $150M zombie module that nobody on the earnings call wants to talk about.

Related: [q1683 — Should Datadog acquire Sentry?](/lab/cheap-100/q1683.json) · [q1685 — Datadog vs New Relic in 2026](/lab/cheap-100/q1685.json) · [q1694 — Should Datadog kill its CI Visibility module?](/lab/cheap-100/q1694.json)