Elastic

We will discontinue analyst coverage of Elastic on or about Sept. 25. 

We provide analyst research and ratings on over 1,500 companies globally and periodically adjust our coverage according to investor interest and staffing.

Our $62 fair value estimate implies an enterprise value of 2.7 times our fiscal 2027 sales estimate, which is lower than other observability peers.

The main drivers of Elastic’s revenue trajectory are artificial intelligence workflows and related search demand, growth in unstructured data, enterprise demand for efficiency-enhancing software, Elastic’s ability to penetrate these markets, and the possibility of competition from AI lab integrations.

With 480 million terabytes of data generated daily (and increasing), roughly 70% of organizations already using AI in at least one business function, it's clear that Elastic is in a fast-growing market, but we believe there are disruption risks should AI labs be capable of direct-to-enterprise backend integrations. Should AI labs' agentic integrations be viable, Elastic's dedicated search and retrieval layer could be commoditized.

To develop our revenue forecasts, we estimate Elastic’s penetration in AI search, observability, and security markets by compiling and analyzing various market-size estimates. We then create a baseline projection based on the attractiveness of the Elastic offerings. This results in an average annual growth rate of 14% over three years, decreasing to 7% by 2036. Overall, this growth path reflects our view that Elasticsearch is a viable AI product today, but will likely face increasing competition for leading AI labs in the future.

Cost of revenue and R&D are the largest expenses, primarily comprising salaries. Cost of revenue also includes third-party cloud hosting costs. Cost of revenue has slightly decreased from 29% of revenue in 2019 to 24% in fiscal 2026, and we project it will compress to 21% by 2036, thanks to Elastic’s advanced data indexing, orchestration, and compression techniques that reduce cloud computing needs for data querying. We expect ongoing operating leverage and computing efficiency improvements throughout our forecast, leading to an estimated gross margin of 79% by 2036—placing Elastic near the top quartile of 309 software companies since 2015. Research and development as a percentage of revenue decreased from 37% in 2019 to 26% in 2026. Although we expect R&D spending in dollar terms to rise, its relatively fixed salary component should scale well as the company expands, leading us to estimate R&D at 20% of revenue by 2035.

Overall, we think Elastic has a strong AI-search product suite and attractive observability and security offerings, but the competition in these three vectors is intense, which should keep forward expectations in check.

We believe Elastic lacks a durable competitive advantage and does not have a moat.

The AI-search segment’s vectorized and inverted indexing capabilities, while technologically impressive, are likely to be challenged by the rise of agentic workflows that are increasingly capable of enterprise-level backend integrations. The observability and security segments, which we estimate make up the remaining two-thirds of the business, lack maintainable competitive advantages because of the many available alternatives in a crowded market.

While we typically see observability and security as infrastructural layers in technology stacks filled with chokepoints that still hold value in an AI-first world, the reality is that competition in these layers is intense, and AI-search remains a significant part of the business. We believe that AI-search’s main functions of indexing, retrieval, and generation are component-like point solutions in nature and can be integrated into products developed by AI labs or bundled by larger platforms.

Furthermore, while the open-source nature and widespread adoption of the Elasticsearch AI-search solution have fostered an ecosystem, we see this primarily as a go-to-market advantage rather than a durable network effect. In reality, we believe that Elasticsearch’s open-source approach has, in some ways, hindered future returns because its accessible codebase has enabled Amazon to "fork" its logic and develop OpenSearch. Consequently, if an Amazon Web Services cloud customer needs a modular indexing and retrieval tool, they may turn to OpenSearch, which can offer most of Elastic's features, limiting cross-selling and upselling opportunities within the broader Elastic suite.

Besides hyperscaler competition through bundled indexing and retrieval, we also see AI-lab-created backend integrations as a major threat to stand-alone platforms like Elastic, which ultimately shortens the runway for future excess returns in search. Recently, we've observed an influx of retrieval tools from Anthropic, Google, and OpenAI that feature native backend integrations via application programming interfaces, or APIs, that can bypass traditional software layers. All three AI labs now enable some form of agent-native retrieval within live enterprise data systems in real time. When agents can query live data systems with governed permissions, it makes dedicated indexing layers like Elastic less essential for enterprise use cases. As a result, we see major risks to the durability of Elasticsearch’s value proposition, and thus, major risks to future excess returns.

Although there is some argument for switching costs due to the mission-critical and infrastructural nature of issues addressed by observability and security, we believe these arguments are overstated for Elastic, especially considering many viable alternatives from observability pure plays and its less-than-stellar relative retention and growth metrics. Compared with Datadog, a leader in the observability industry, Elastic lags behind in net revenue retention by 17 percentile points (75th percentile versus 92nd percentile) within a universe of 34 software companies. In revenue growth, Elastic falls short relative to Datadog by nearly 30 percentile points (60th percentile versus 90th percentile). Elastic compares similarly on these dimensions with narrow-moat Dynatrace, but we believe Dynatrace’s core clientele of the largest enterprises, with the most complicated technology infrastructures, is stickier. Overall, Elastic operates in markets where customers have many alternatives, including hyperscaler-native tools, AI-lab tools, and top-tier pure plays, which ultimately reduces the likelihood of maintainable excess returns above the cost of capital.

Elastic’s combination of vectorized search and inverted indexes for keyword search is fast and exhibits contextual accuracy, which can enable it to stand out from other search solutions.

Elastic’s cloud-neutral approach makes it accessible to everyone and reduces hyperscaler lock-in, an attractive value proposition to the majority of firms that prefer multicloud services.

Elastic’s data compression techniques, node orchestration, and auto-scaling of servers allow customers to easily set and forget a variety of complex parameters.

AI labs have been rapidly innovating their agentic products. If agents were integrated directly from the AI lab to the enterprise, it could commoditize the Elasticsearch solution.

Open-source availability of Elastic’s core features will cannibalize paid seat expansion opportunities.

Best-of-breed bakeoffs favor formidable competition from observability and security pure plays like Datadog and CrowdStrike.