Dynatrace, Inc.
DT 6.5Price and target zones价格与目标区间
What the company does公司做什么
Dynatrace makes software that watches a company's applications and computers and automatically finds what is wrong when something breaks or slows down. Instead of dashboards a person must read, its AI points straight to the cause.Dynatrace 制造的软件，会盯着一家公司的应用程序和电脑，在出故障或变慢时自动找出问题所在。它不用人去盯一堆仪表盘，而是靠AI直接指向病根。
Earns from营收来源 subscriptions. In FY2025, 95% of the $1.7 billion revenue was subscription, and recurring revenue (ARR) was $1.7 billion, up 15%.订阅。2025财年，17亿美元收入中有95%来自订阅，年度经常性收入（ARR）为17亿美元，增长15%。
Read the full description查看完整介绍
Dynatrace, Inc. engages in the advancement of observability for digital businesses, which transforms the complexity of modern digital ecosystems in North America, Europe, the Middle East, Africa, the Asia Pacific, and Latin America. The company operates Dynatrace, an AI-powered observability platform, which provides solutions, including infrastructure, application, threat, and AI observability; digital experience; log analytics; application security; software delivery; and business analytics. It also offers implementation, consulting, and training services. The company markets its products through a combination of global direct sales team and a network of partners, including global system integrators (GSIs), cloud providers, resellers and technology alliance partners. It serves customers in various industries, including banking, financial services, government, insurance, retail and wholesale, transportation, and software. Dynatrace, Inc. was founded in 2005 and is headquartered in Boston, Massachusetts.
Revenue breakdown营收构成 share of revenue营收占比
This is Dynatrace's core platform, sold as a subscription that customers renew every year. It covers application and infrastructure monitoring, log management, and an AI engine that finds the root cause of a problem automatically, instead of leaving a person to read charts. Because the platform embeds itself deep into a customer's systems, renewal rates stay high and revenue is predictable.这是Dynatrace的核心平台，以订阅方式销售，客户每年续订。它涵盖应用与基础设施监控、日志管理，以及能自动找出问题根本原因的AI引擎，让人无需再靠读图表来诊断问题。由于该平台深度嵌入客户的系统，续约率保持较高，收入也更可预测。
- Application and infrastructure monitoring应用和基础设施监控 — Watches software and servers across clouds and traces problems end to end.跨多个云监控软件和服务器，并端到端追踪问题。
- Log management日志管理 — Collects and analyses the records that systems produce.收集并分析系统产生的记录。
- Davis AI — Works out the root cause of a problem automatically, rather than leaving a person to read charts.自动找出问题的根本原因，而不是让人去看图表。
This is fee-based work to set up and configure the platform for new customers, plus ongoing consulting. It is a small, largely one-time complement to the subscription business rather than a growth driver, and its share of revenue shrinks as the subscription base scales.这部分是为新客户搭建和配置平台的收费工作，加上持续的咨询服务。这是订阅业务的一小部分、多为一次性的补充，而非增长驱动力，随着订阅业务规模扩大，其收入占比会持续缩小。
Company facts公司资料
- Founded成立时间
- 2005
- Headquarters总部
- Boston, Massachusetts
- Employees员工人数
- 5,600
- Industry行业
- Software - Application
- Country国家
- United States
Investment thesis投资逻辑 from the watchlist来自关注列表
🟡 ANALYZED 8/06 — first file ever, first conviction score. HOLD. Both withdrawn zones re-derived. Entry $38-43 FV $42-56 (central ~$49) Trim 27x fwd (renders ~$61).
What's working
- Q1 FY27 beat every guided line and guidance was RAISED operationally (cc revenue +25bps, EPS +$0.04, margin high end +25bps) — every reported-dollar cut is 100% FX. Logs ~$200M annualized, doubled in 2 quarters; AI-cohort customers consume 1.5x more
- Offsetting, and it is the tail risk: there is NO churn crisis — exactly one public displacement exists anywhere (anonymized, ~20% cost saving), gross retention holds mid-90s, and vs Datadog DT repeatedly wins on price. The bear case is expansion compression, not churn. Conviction at 6.5 — the platform wound explains data already penalized (flat logos, 110% NRR); do not double-count a mechanism whose effect is already in the numbers.
Red flags
- THE REPORTED EARNINGS SERIES IS VOID: DT released a deferred-tax valuation allowance in FY25 — tax provision −$260.3M (benefit), ~$0.86 of the $1.59 EPS — then ran a 45.7% rate in FY26. The 89.5x trailing P/E, the −25% "earnings growth", and ROE are all artifacts → [[pitfall-tax-valuation-allowance-round-trip-breaks-eps]]. Normalized at 21% tax, pretax income GREW 34% in FY26 ($223M→$300M) while GAAP EPS "fell" 66%; operating income CAGR +41.7%/3yr, FCF/share +15.1%/3yr, debt/assets 3.7%
- New: the 2nd→3rd gen platform migration is a live product wound — "almost akin to a SAP implementation," "I will definitely recommend people not to buy this product at this time" — and Dynatrace's own DevRel publicly conceded the cause. Still generating complaints in Jan-2026 reviews; a plausible driver of the flat logo count. 🔍 HN mindshare since 1/2025: Datadog 3,131 Grafana 2,315 New Relic 82 DT 28 — one-third of New Relic's
- $478.7M of FY26 buybacks cut the DILUTED share count by ZERO (303.60M→303.73M) — the whole buyback offsets SBC, which also guts the Starboard ">$2.5B returned" catalyst
- CFO Benson retires by 3/31/27, no successor
Watch
- But the moat is 5.0/10 with a NEGATIVE derivative: customer count FLAT at ~4,100 for two years while R&D intensity rose 470bps, NRR pinned at 110% for three quarters (was 119%), and OTel graduated the CNCF 5/11 — standardizing away the OneAgent lock-in both top moat sources depend on. 🔻 Practitioner-evidence pass closed 8/06 → switching costs cut 6.5 → 5.5: across Reddit/HN/PeerSpot/Capterra, NOBODY frames OneAgent as lock-in — it is praised as "easy mode for telemetry," i.e. a time-to-value moat, exactly what OTel erodes.
- Yahoo's 21.4x fwd P/E uses FY28 EPS ($2.26) not the guided FY27 $1.98 → real multiple 24.4x → [[pitfall-vendor-forward-eps-is-the-wrong-fiscal-year]]; consensus tgt $47.91 is stale (six firms went to $59-62 on 8/05-06). ⏳ THE WHOLE THESIS IS THE 2H FY27 DPS RENEWAL COHORT — ~70% of renewals, all three cohorts resetting; NRR either inflects past 110% or the growth case breaks. Do not add at fair value into 20+ other software names.
- Sleeve question flagged, not actioned: a 5.5/10 decaying moat arguably fails this sleeve's "hold through a 40% drawdown without re-underwriting" test.
🟡 8/06 首次分析 — 首个文件、首个信心评分。 持有不动。 两个被撤回的区间已重新推导。 ** 入场 $38-43 合理价值 $42-56（中枢约 $49）· 减持 27x fwd。 ** 🚩 申报的每股收益序列是无效的： DT 在 FY25 释放了递延税项估值备抵 — 税项 −$260.3M（收益），占 $1.59 每股收益中的约 $0.86 — 随后在 FY26 录得 45.7% 的税率。 89.5x 的滚动市盈率、−25% 的「盈利增长」和净资产收益率全是假象 → [[pitfall-tax-valuation-allowance-round-trip-breaks-eps]]。 按 21% 税率还原后，FY26 税前利润增长 34%（$223M→$300M），而 GAAP 每股收益却「下跌」66%；营业利润 3 年复合增长率 +41.7%，每股自由现金流 +15.1%，负债/资产 3.7%。 ✅ FY27 第一季度超出每一条指引，且 指引在经营层面是上调的（固定汇率收入 +25bps、每股收益 +$0.04、利润率上限 +25bps）— 所有以美元计的下调 100% 来自汇率。 日志年化消费约 $200M，两个季度翻倍；AI 客户群的消费量高出 1.5 倍。 🔴 但护城河为 5.0/10 且导数为负：客户数量已连续两年持平在约 4,100 家，而研发强度上升 470bps；净留存率连续三个季度钉在 110%（曾为 119%）；OTel 已于 5/11 从 CNCF 毕业 — 正在把 OneAgent 锁定效应标准化掉，而这正是两大护城河来源所依赖的东西。 ** 🔻 8/06 完成一线从业者证据补充 → 转换成本由 6.5 下调至 5.5：在 Reddit/HN/PeerSpot/Capterra 上，没有任何人把 OneAgent 描述为锁定 — 它被称赞为「遥测数据的简单模式」，也就是说这是一条 价值实现速度 的护城河，而这恰恰是 OTel 正在侵蚀的东西。 🚩 新发现：第二代到第三代平台的迁移是一处尚未愈合的产品伤口 — 「几乎相当于一次 SAP 实施」、「我一定会建议大家现在不要买这个产品」 — 而且 Dynatrace 自己的开发者关系人员公开承认了原因。 到 2026 年 1 月的评论中仍在产生抱怨；这很可能是客户数量停滞的一个原因。 🔍 自 2025 年 1 月以来的 HN 讨论量：Datadog 3,131 Grafana 2,315 New Relic 82 DT 28 — 只有 New Relic 的三分之一。 ✅ 反向证据，而且这正是尾部风险所在：不存在客户流失危机 — 全网仅有一例公开的替换案例（匿名，成本仅节省约 20%），总留存率维持在 90% 中段，且 在价格上 DT 屡次胜过 Datadog。 ** 空头逻辑是扩张放缓，而非客户流失。 信心评分维持 6.5 — 平台伤口解释的是已经被扣分的数据（客户数持平、净留存率 110%），不应对同一机制重复扣分。 ** 🚩 FY26 的 $478.7M 回购把稀释后股数降低了零（303.60M→303.73M）— 整笔回购都在抵消股权激励，这也架空了 Starboard「三年返还 >$2.5B」的催化剂。 🚩 首席财务官 Benson 于 2027/3/31 前退休，尚无继任者。 ⚠️ 雅虎 21.4x 的前瞻市盈率用的是 FY28 每股收益（$2.26），而不是指引中的 FY27 $1.98 → 真实倍数为 24.4x → [[pitfall-vendor-forward-eps-is-the-wrong-fiscal-year]]；一致目标价 $47.91 已 过时（8/05-06 有六家机构上调至 $59-62）。 ⏳ 整个投资逻辑就是 FY27 下半年的 DPS 续约批次 — 约 70% 的续约、三个批次全部首次重置；净留存率要么突破 110%，要么增长逻辑破裂。 Benson 已预先承诺在约 11 月 5 日的财报中对全年 ARR 指引作年中修订。 *不要在合理估值处、在已有 20 多只软件股的组合里加仓。 ** ⚠️ 已标记但未执行的分类问题：一条 5.5/10 且正在衰减的护城河，恐怕通不过本篮子「在 40% 回撤中持有而无需重新论证」的检验。 *
Historical results历史业绩 past fiscal years — these values do not change过去财年——这些数值不会变动
Profit margins利润率 the share of revenue that stays as profit营收中留存为利润的比例
of every $1 of revenue每 $1 营收中的占比
- Return on equity净资产收益率
- 6%
- Revenue growth营收增长
- +16%
Competitors竞争对手
Why customers stay客户为什么留下
One software agent plus an AI engine that pinpoints the cause, instead of leaving a human to diagnose it. The more of a customer's systems it watches, the deeper the data and the higher the switching cost.一个软件探针加上一套AI引擎，能精准找出病根，而不用人去做诊断。它盯着客户的系统越多，积累的数据就越深，客户的转换成本也就越高。