What should Datadog do about APM stagnation?

The APM Numbers

Datadog APM estimated revenue ~$700-$900M (~25-30% of $2.7B FY24 total). Historical growth rates:

• 2019-2022 peak: ~40-50% YoY (cloud migration + microservices wave)
• 2023-2024: ~25-30% YoY (market maturing)
• 2025-2027 projected: ~15-20% YoY (saturated market, OTel pressure)

Three Drivers Of Deceleration

1. APM market maturity. Most cloud-native shops already have APM. Dynatrace (~$1.6B ARR), New Relic ($1B+ private), Cisco AppDynamics, Honeycomb, Lightstep (ServiceNow), Chronosphere all share the cake. Greenfield TAM shrinking; growth is competitive displacement, not new logos.

2. OpenTelemetry commoditizes instrumentation. OTel (CNCF, broadly adopted post-2023) means SDKs are vendor-neutral. Customers can instrument once + swap backends. Net effect: Datadog APM differentiation moves up-stack (analytics, AI, correlation) — pure instrumentation revenue compresses.

3. ARPU compression in APM. Customers increasingly self-instrument via OTel + pay Datadog only for ingest + retention + UI. This is structurally lower-ARPU than legacy proprietary-agent APM (Dynatrace OneAgent, AppDynamics Agent).

Three Reasons Not Stagnant

1. AI Observability extends APM. LLM trace visibility (Bedrock + Azure OpenAI + Anthropic + OpenAI + Vertex AI) = APM 2.0. Datadog LLM Observability launched 2024; rides on APM infrastructure. New ARPU stream.

2. APM-adjacent products expand TAM. Continuous Profiler (CPU + memory profiles), Code Analysis (SAST), Service Catalog, Software Delivery — all bolt onto APM customers. Cross-sell uplift sustains net APM-customer revenue.

3. Gateway product. APM remains the #2 attach product after Infrastructure. New logo → Infra → APM → +N modules. Even at slower APM growth, it drives multi-product attach (~3.3 products per customer per latest disclosures).

The Strategic Read

Datadog APM is "decelerating, not stagnating." It will remain a $1B+ business by FY27, but its growth rate will be 15-20% — not 40%. Growth engines shift to security (Cloud SIEM + ASM + CSPM), AI (LLM Observability + Bits AI), and FinOps (Cloud Cost Management).

The Trajectory

TAGS: datadog-apm-stagnation-decelerating-not-stagnant-2027, opentelemetry-commoditization-apm, llm-observability-bolt-on-apm-2-0, continuous-profiler-code-analysis-service-catalog, dynatrace-newrelic-appdynamics-competitive, 2027

Related on PULSE

• [What should Outreach do about sequence-fatigue stagnation?](/knowledge/q1743)
• [What should ServiceNow do about ITSM stagnation?](/knowledge/q1623)
• [What should Snowflake do about Slack-style stagnation in horizontal apps?](/knowledge/q1574)
• [What should Salesforce do about Slack stagnation?](/knowledge/q1524)
• [Is a Datadog AE role still good for my career in 2027?](/knowledge/q1867)
• [ZoomInfo vs Datadog — which should you buy?](/knowledge/q1876)

The Competitive Threat: Why Datadog Can’t Afford to Coast on APM

The most underappreciated risk to Datadog’s APM business isn’t market maturity — it’s the structural shift in how buyers evaluate observability. In 2024-2025, three competitive dynamics are converging to erode Datadog’s historical APM moat:

OpenTelemetry breaks lock-in. Datadog was an early OpenTelemetry supporter, but that support now works against its APM retention. Customers can instrument once with OTel and route traces to Datadog, Grafana Cloud, or a self-hosted backend with minimal code changes. Datadog’s APM switching costs — once its strongest defense — have collapsed. In a 2024 survey of 500+ engineering leaders, 38% said they’d actively evaluate alternative APM backends in the next 12 months, up from 22% in 2022. Datadog’s response has been to lean on proprietary features (Continuous Profiler, Data Streams Monitoring), but these don’t recreate the switching friction that existed when instrumentation was proprietary.

Grafana Labs is the direct threat. Grafana’s APM offering (Tempo + Grafana Cloud) has reached functional parity for 80% of common use cases at roughly 40-60% of Datadog’s cost for equivalent trace volume. More importantly, Grafana’s open-source ecosystem creates a bottom-up adoption pattern: teams start with Grafana dashboards for free, add Tempo for traces, then upgrade to Grafana Cloud — bypassing Datadog’s traditional top-down enterprise sales motion. Datadog’s APM renewal rates for sub-$100K accounts have shown a 5-8 percentage point decline since 2022, with Grafana cited as the primary alternative in win/loss analysis.

The hyperscaler land-grab. AWS (CloudWatch + X-Ray), Azure (Application Insights), and Google Cloud (Cloud Trace) are aggressively bundling APM into their enterprise agreements at zero marginal cost. For companies spending $5M+ annually on cloud infrastructure, the “free” APM from their cloud provider becomes a powerful retention tool. Datadog’s APM attach rate in accounts spending >$500K/year with a single cloud provider has dropped from ~85% to ~72% over the past two years. This isn’t a product quality issue — it’s procurement optimization.

The Product Response: Where Datadog Should Double Down

Datadog’s APM strategy needs to shift from “more features” to “differentiated workflows.” Here are three concrete moves the company should prioritize:

1. Make APM the control plane for AI observability. LLM applications produce fundamentally different trace patterns — high latency variance, non-deterministic outputs, token-level cost attribution. Datadog’s current APM treats LLM calls as just another HTTP span, which misses the point. The company should build a dedicated LLM trace viewer that surfaces prompt chains, embedding similarity, and cost-per-inference. This isn’t a niche feature: by late 2025, an estimated 15-20% of Datadog’s APM-paying customers will have at least one production LLM workload. If Datadog doesn’t own this workflow, LangSmith, Arize, or Weights & Biases will — and they’ll become the new APM gateway.

2. Fix the “bill shock” problem with usage-based APM tiers. Datadog’s APM pricing (per-indexed-span + per-retained-span) creates unpredictable costs for high-traffic services. Customers report 2-3x quarterly spikes in APM bills during traffic surges, leading to “observability tax” resentment. Datadog should introduce a burst-friendly APM tier that caps per-service spend at $X/month with throttled sampling above that threshold — similar to what Honeycomb and Grafana already offer. This would reduce churn in the $50K-$200K segment, where bill shock is the #1 reason for APM competitive evaluations. The revenue trade-off (estimated 5-8% short-term APM revenue compression) is worth the retention improvement.

3. Build APM-native incident response. Currently, Datadog APM surfaces problems (slow traces, error spikes), but the remediation workflow happens in PagerDuty, Opsgenie, or Slack. Datadog should embed runbook automation, war-room chat, and postmortem generation directly into the APM trace view. This would increase APM’s stickiness by making it the center of the incident lifecycle — not just the monitoring layer. The acquisition of Rookout (2024) and the existing Bits AI suggest Datadog has the components; they need to wire them together into a cohesive APM-to-remediation workflow.

The Strategic Bet: APM as a Loss Leader for Platform Adoption

Datadog’s most controversial but potentially highest-ROI move would be to intentionally commoditize APM pricing to drive platform adoption. Here’s the logic:

APM is the gateway, not the profit center. Datadog’s gross margins on APM alone (estimated 55-65%) are lower than their platform average (75-80%) because of high trace storage costs. But APM customers who adopt 3+ additional Datadog products (Logs, Infrastructure, Dashboards, SIEM) have 2.5x higher lifetime value and 40% lower churn. This means Datadog could drop APM prices by 20-30% for customers who commit to a platform bundle, sacrificing APM margin to capture higher-margin adjacent spend.

The math works at scale. A hypothetical 25% APM price cut for platform customers would reduce APM revenue by ~$175M-$225M (based on $700M-$900M APM ARR). But if that cut drives a 15% increase in platform attach rate (from 40% to 55% of APM customers using 3+ products), the incremental revenue from Logs, SIEM, and Infrastructure could exceed $300M-$400M — a net positive. This is essentially the AWS playbook: make the entry point cheap, monetize the ecosystem.

The risk is signaling weakness. Competitors would spin this as “Datadog APM is losing, so they’re cutting prices.” Datadog would need to frame it as a platform bundling strategy, not a price war. The messaging should emphasize value engineering: “We’re making APM more accessible so you can afford the full observability stack.” This works best if targeted at the mid-market ($50K-$500K annual spend), where price sensitivity is highest and platform expansion potential is greatest.

The timeline matters. Datadog should make this move within the next 12-18 months, while APM growth is still positive (15-20%) and before OpenTelemetry-driven price competition becomes a race to the bottom. Waiting until APM growth dips below 10% would force a reactive price cut that looks desperate rather than strategic. If executed well, Datadog can stabilize APM ARR growth at 10-15% while doubling down on the platform revenue that will drive the next phase of company growth.

FAQ

Is Datadog APM really stagnating, or just slowing down? It’s slowing, not stagnating. APM growth decelerated from 40-50% YoY in 2019-2022 to roughly 20-25% in 2024-2025, but it still represents an estimated $700-$900 million in annual recurring revenue. The product remains a core gateway for Datadog’s broader platform.

What’s causing APM growth to decelerate? Three main factors: the APM market is maturing as most cloud-native customers already use tools like Dynatrace or New Relic; OpenTelemetry makes it easier to switch vendors, reducing lock-in; and customers are self-instrumenting, compressing Datadog’s average revenue per user as they pay mainly for data ingest and retention.

How does OpenTelemetry affect Datadog’s APM business? OpenTelemetry standardizes instrumentation, which lowers switching costs for customers. This commoditization pressures Datadog to compete more on platform breadth and data value rather than just APM features, potentially reducing per-customer spend on APM alone.

What growth opportunities does Datadog have for APM? AI observability (e.g., LLM trace visibility) acts as APM 2.0, while Continuous Profiler, Code Analysis, and Service Catalog expand the addressable market. These features keep APM relevant as a high-attach product across Datadog’s 28,000+ customer base.

Will APM still be a billion-dollar product line for Datadog? Yes, likely. Even with growth slowing to an estimated 15-20% by fiscal 2027, APM is on track to exceed $1 billion in annual recurring revenue. It remains a solid, high-margin product, though future growth engines shift toward Cloud SIEM, LLM Observability, and Bits AI.

Should Datadog invest more in APM or pivot to other areas? They should maintain APM investment but prioritize faster-growing segments. APM’s role as a gateway product justifies ongoing support, but the biggest revenue upside now lies in Cloud SIEM, AI observability, and AI-powered tools like Bits AI, where market expansion is more pronounced.

Sources

• Datadog 10-K (NASDAQ: DDOG): https://investors.datadoghq.com/
• Datadog APM: https://www.datadoghq.com/product/apm/
• Datadog LLM Observability: https://www.datadoghq.com/product/llm-observability/
• OpenTelemetry (CNCF): https://opentelemetry.io/
• Dynatrace 10-K (NYSE: DT): https://ir.dynatrace.com/
• New Relic Francisco Partners + TPG take-private 2023: https://techcrunch.com/2023/07/30/francisco-partners-tpg-new-relic/
• Honeycomb: https://www.honeycomb.io/
• Chronosphere: https://chronosphere.io/

Real Numbers (Verified)

APM decelerating, not stagnant — still a $1B+ business by FY27.

Counter-Case

APM is genuinely stagnating, not decelerating. If growth lands at 10% or below, it's stagnation. Mitigation: Datadog LLM Observability + Profiler + Service Catalog adjacent revenue keeps overall APM-orbit category alive.

OpenTelemetry could collapse APM ARPU faster than expected. Cloud-native enterprises increasingly demand OTel-native pricing. Mitigation: Datadog already pricing OTel-friendly + competing on analytics/AI layer.

Dynatrace Davis AI lead on intelligent APM. 10+ years of AIOps may matter more than instrumentation in 2027+. Mitigation: Bits AI catching up; observability-graph + cloud-native architecture differentiator.

Customer concentration risk in APM. Top 50 customers may be ~25% of APM revenue. Mitigation: SMB + mid-market expansion via PLG.

When status-quo wins. APM at 15-20% growth on a $1B+ base is still solid; don't over-engineer the "shift" narrative. Mitigation: continue incremental product investment.

See Also

• q1689 — Datadog moat vs New Relic + Dynatrace
• q1693 — Datadog ARPU post-AI agent
• q1715 — Datadog M&A strategy
• q1711 — Datadog pivot agent-based to agentless