Three Cloud Observability Stocks Trading on Different Trajectories

Datadog generated $3.4 billion of revenue in 2025, growing 28% year-over-year. The growth rate has decelerated from the 60%+ peak of 2021-2022 but the absolute revenue base is now substantial enough that the deceleration is mathematically inevitable. The free cash flow line tells the more interesting story; $1.0 billion of FCF in 2025 against $836 million in 2024, a 20% expansion. The FCF margin sits at 29%, comfortably the highest in the observability peer group. The operational signature reflects the platform model where customer expansion across modules drives high-incremental-margin revenue.

The AI monitoring product has been the standout growth driver in the trailing twelve months. The LLM Observability product, launched in mid-2024, has scaled to over $400 million annualised contribution by Q4 2025. Customers have integrated the AI monitoring tools as part of broader enterprise AI deployment workflows, and the customer cohort that has adopted AI monitoring has shown net dollar retention rates north of 130%. The integration with the broader Datadog platform produces meaningful upsell across the existing customer base.

The enterprise customer mix has continued to expand. Datadog ended 2025 with approximately 4,200 customers spending over $100,000 annually, up 18% year-over-year. The largest customers (those spending over $1 million annually) crossed 540, up 23% year-over-year. The customer expansion at the high tier is the operational moat; once a customer has integrated 8-10 Datadog modules into the production environment, the switching cost becomes meaningful and the renewal economics become predictable.

Dynatrace generated approximately $1.7 billion of revenue in fiscal 2025 (year ending March), growing 19% year-over-year. The growth rate is below Datadog's but the customer mix is meaningfully different. Approximately 75% of Dynatrace's revenue comes from large enterprise customers, against approximately 60% at Datadog. The enterprise mix produces a more predictable revenue trajectory but slower top-line acceleration. Operating margins sit at roughly 27% on a non-GAAP basis, which is competitive with Datadog at scale.

The platform consolidation thesis has been the operational tailwind for Dynatrace. As enterprises rationalise their observability tool stacks, the broader platform vendors (Dynatrace, Datadog, Splunk via Cisco) gain at the expense of the point-solution providers. Dynatrace's all-in-one architecture has been particularly well-suited to the consolidation buying pattern, with the average customer deploying 6-8 Dynatrace modules against an industry average of 3-4. The consolidation play has another 18-24 months of tailwind to run.

The AI-powered analytics features (Davis AI) have been the differentiator at the enterprise tier. Davis automates the root-cause analysis workflow that historically required engineering team time, producing measurable productivity gains for the customers who deploy it. Davis-related revenue contribution is approaching 15% of total ARR by Q4 2025 and is growing at 35-40% annually. The product roadmap suggests continued integration with broader enterprise AI workflows, which should sustain the differentiation.

The third leg of the observability sector is populated by the smaller and the recently-acquired names. New Relic was taken private by Francisco Partners and TPG in 2023 and is no longer publicly tradable. Splunk was acquired by Cisco in early 2024. The rump of independent smaller-cap observability names is now small and the public market exposure to the sector is concentrated in the two larger pure-plays.

The Splunk-Cisco combination is the structural feature worth flagging. Cisco's enterprise networking footprint creates a distribution channel for the integrated Splunk observability platform that the pure-plays cannot easily match. The competitive dynamic has shifted in the trailing twelve months as Cisco's go-to-market begins to land in the enterprise tier. Datadog and Dynatrace have responded with deeper integration partnerships at the cloud hyperscaler level (AWS, Azure, GCP), which provides a counterbalancing distribution channel. The competitive equilibrium is recalibrating but neither pure-play has been displaced.

The broader sector dynamic is that the observability category continues to consolidate around two or three large platform vendors. The mid-tier and the application-specific monitoring tools are losing share to the integrated platforms. This consolidation is the structural tailwind that supports both Datadog and Dynatrace, although in different ways. Datadog captures more of the modern microservices-architected enterprise customer; Dynatrace captures more of the traditional enterprise migrating from legacy monitoring tools.

Net dollar retention is the operational metric that best predicts the next-twelve-months revenue trajectory in software platform businesses. Datadog reported net dollar retention of approximately 116% in Q4 2025, down from 121% the prior year. The deceleration reflects the optimisation behaviour at the largest customers, who have been actively reducing the volume of monitored data through sampling, retention tuning, and architecture changes. The 116% figure is still healthy but the deceleration trajectory is real.

Dynatrace reported a comparable retention metric, the net retention rate, of approximately 111% on a constant currency basis. The lower headline number reflects both the larger enterprise customer mix (large customers tend to have flatter retention than mid-market customers) and the slower modular expansion at the all-in-one platform compared to Datadog's a la carte module model. The trajectory at Dynatrace has been more stable over the trailing four quarters, with the net retention rate holding in the 110-113% range rather than decelerating.

The forward read on retention is the structural feature that supports the differential calls. Datadog's revenue trajectory is more sensitive to customer optimisation behaviour because the consumption-based pricing model amplifies the optimisation impact. Dynatrace's predominantly subscription-based model insulates the revenue line from short-term optimisation cycles. In a softening enterprise IT spending environment, Dynatrace's revenue trajectory should prove more resilient. In an accelerating environment, Datadog should over-perform. The current macro setup is closer to the former than the latter, which supports the relative call to overweight Dynatrace.

The historical analogue is the Salesforce-Workday comparison through the 2015-2017 enterprise spending pause. Workday's subscription-heavy revenue model proved more resilient through the spending pause, while Salesforce's broader product mix re-accelerated faster when enterprise IT spending recovered. The current observability cycle is at the first half of the analogous pattern. The Signals Desk read is to position around that historical sequencing.

AI monitoring has emerged as the most consequential product category extension in the observability sector in the past three years. As enterprises deploy generative AI applications into production, the operational requirements for monitoring LLM behaviour, tracking inference latency, observing prompt and completion patterns, and detecting hallucination drift have created a new category of monitoring requirements that the traditional APM stack cannot address. Both Datadog and Dynatrace have built dedicated AI monitoring product lines.

Datadog's LLM Observability product, launched in mid-2024, has scaled to over $400 million annualised contribution by Q4 2025. The product integrates with the major LLM providers (OpenAI, Anthropic, Cohere) and provides span-level tracing of LLM-augmented applications. The customer cohort that has adopted LLM Observability has been growing at over 200% annually, with the average ARR contribution per customer expanding as deployment matures. The product fits naturally into the consumption-based pricing model and contributes to the platform's land-and-expand motion.

Dynatrace's AI monitoring suite is anchored on the Davis AI platform, which combines automated root-cause analysis with the broader application performance monitoring stack. Davis-related ARR is approaching 15% of total ARR by Q4 2025 and is growing at 35-40% annually. The Davis differentiation is the integration with the broader enterprise observability workflow rather than a standalone LLM monitoring product. Both approaches have merit; Datadog's product is purer-play AI monitoring, while Dynatrace's is more integrated with traditional enterprise IT operations.

The sector revenue contribution from AI monitoring is on track to scale from approximately $700 million in 2025 to $2.5-3.0 billion by 2027. Both platforms are well-positioned to capture meaningful share of that expansion. The product roadmaps at both companies indicate continued investment in the category. The competitive dynamic between the two products will be a multi-year story, with neither side likely to dominate definitively.

The portfolio implication for an observability sector allocation is to weight Dynatrace at roughly 55-60% of the sector basket, Datadog at 35-40%, and to leave a smaller portion for selective exposure to the broader enterprise software stack via the integrated platforms. Dynatrace's near-term risk-reward favours overweight positioning. Datadog's longer-term FCF profile and AI monitoring leadership justify the core allocation. The sector is not a binary choice between the two names; it is a relative weighting decision.

The entry zones for both names support measured accumulation rather than chase buying. Datadog at $144 against fair value of $145-165 implies modest upside but the volatility profile (beta 1.29) suggests the share price will trade through the 200-day moving average ($138) at multiple points over the next 12 months. Adding on those pullbacks rather than at the current price improves the risk-adjusted entry. Dynatrace at current levels offers the cleaner upside to fair value of $70-82.

The sector-level catalyst path is the next two earnings cycles, the AI monitoring revenue continued scaling, and the enterprise IT spending trajectory. Each of those data points provides incremental confirmation or refutation of the structural growth thesis. Across two complete software platform cycles, the pattern at this point in the consolidation curve has favoured the second-tier valued name (Dynatrace) over the premium-multiple leader (Datadog) on a 12-month risk-adjusted return basis. The setup is repeating with conviction.