# Amplitude, Inc. (AMPL) — Investment Thesis

**Exchange:** Nasdaq  
**Coverage as of:** 2026-Q2  
**Updated:** 2026-06-17  
**Tier:** Free primer (steps 1 & 3 of 19)  
**Sibling pages:** /stocks/ampl/financials · /memo/ampl

> This page shows the free thesis context (business model + recent catalysts).
> The full investment thesis (moat analysis, DCF, scenarios, risk register) is available
> via GET /api/v1/research/AMPL/memo ($2.00, Bearer token).

## Business Model

---
source: coverage-next-full
ticker: AMPL
company: Amplitude, Inc.
step: 01
title: Business Overview
date: 2026-06-14
---

### Step 01 — Business Overview: Amplitude, Inc. (AMPL)

---

#### 1. Business Description

Amplitude, Inc. is a San Francisco-based software-as-a-service company that operates what it calls an AI Analytics Platform — a purpose-built system for collecting, storing, and analyzing behavioral event data generated by users interacting with digital products. [FACT] The company was founded in 2012 by Spenser Skates (CEO), Curtis Liu (CTO), and Jeffrey Wang, went public via direct listing on Nasdaq on September 28, 2021 (ticker: AMPL), and reported FY2025 revenue of $343.2 million (+15% year-over-year) on a fiscal year ending December 31 [S1].

The company's core platform — historically centered on product analytics — answers the question: *who does what, in what sequence, and with what outcome?* It ingests user interaction events (clicks, page views, feature activations, conversions, session starts) via client-side and server-side SDKs, stores them in a proprietary behavioral database called the **Behavioral Graph**, and surfaces the results as funnel analyses, retention charts, user journey maps, cohort comparisons, and AI-generated recommendations. [FACT, S1]

**One-sentence value proposition:** Amplitude enables product, data, and growth teams to understand exactly how users behave inside their digital products — and to act on those insights through experimentation, personalization, and AI-driven automation — replacing a fragmented set of point tools with a single connected platform. [JUDGMENT]

Amplitude's customer base spans SMB, mid-market, and enterprise, across verticals that include B2B SaaS, fintech, e-commerce, media/streaming, and consumer mobile. [FACT] As of December 31, 2025, the company served 4,797 paying customers, including 27 of the Fortune 100, with 698 customers generating $100,000 or more in annualized recurring revenue (ARR) — representing 78% of total ARR of $366 million [S1, S3].

---

#### 2. Value-Chain Layer Map

Amplitude occupies the **analytics and activation layer** in the modern digital product data stack, sitting above the data collection and storage infrastructure and below the decision-execution layer (deployment tooling, CRM, ad platforms). The full stack, from top to bottom, is:

| Layer | What it does | Who owns it |
|---|---|---|
| Data Collection | SDK-level event capture from browser, mobile, server | Amplitude SDKs; also customer's own pipelines (Segment, Rudderstack, Kafka) |
| Data Transport & Ingestion | Event stream delivery to storage | Amplitude-managed (for hosted path); partner CDPs for pre-ingested data |
| Behavioral Storage | Proprietary Behavioral Graph; also native integration with Snowflake/BigQuery/Databricks | Amplitude owns its hosted database; warehouse-native path runs in customer's cloud |
| Analytics Computation | Funnel, retention, cohort, path, impact analysis; ML-powered segmentation | Amplitude (core proprietary value) |
| Visualization & Collaboration | Charts, dashboards, notebooks, AI-generated insights surfaced to analysts, PMs, executives | Amplitude |
| Activation / Decisioning | Feature flags, A/B experiments, audience segments pushed to downstream channels | Amplitude (Feature Experimentation, Audience Activation, Guides & Surveys); partner integrations for downstream CRM/ad tools |

[FACT, S1, S3]

**What Amplitude owns vs. what it partners on:**

- [FACT] Amplitude owns its behavioral data store (the Behavioral Graph), its analytics computation layer, and all product surface area from ingestion to visualization.
- [FACT] Amplitude does **not** require customers to route data through its cloud: its warehouse-native integrations allow customers who store data in Snowflake, BigQuery, or Databricks to query behavioral data directly from those environments, avoiding a separate data copy in Amplitude's cloud. This is a strategic architectural choice designed to serve data-mature enterprise buyers.
- [FACT] Amplitude relies on third-party cloud infrastructure (AWS, GCP) for compute and storage for its hosted cloud product. This creates a variable cost relationship with cloud providers that is reflected in cost-of-revenue growth.
- [JUDGMENT] The warehouse-native architectural bet is Amplitude's most significant structural differentiator vs. earlier-generation analytics tools (Mixpanel, legacy Heap) that require a separate data silo. It positions Amplitude to win deals where Snowflake or BigQuery is already the organizational data source of truth.

---

#### 3. Revenue Model

**Primary revenue stream: subscriptions.** [FACT] Subscription revenue was approximately 98% of total FY2025 revenue of $343.2 million; professional services (implementation, onboarding) accounted for roughly 2%, or approximately $7 million [S1].

**Pricing model:** Amplitude uses a tiered SaaS subscription model with four plans (Free, Plus, Growth, Enterprise). Pricing dimensions have evolved:
- [FACT] Historically, pricing was based on event volume (total events tracked per month). This model created risk for Amplitude when customers with high-volume but low-value event implementations churned or downsized.
- [FACT] Amplitude introduced Monthly Tracked User (MTU) and session-based pricing as an alternative metric, designed to better align cost with business value (engaged users) rather than raw instrumentation volume.
- [FACT] Enterprise accounts are custom-priced through direct negotiation, typically on annual or multi-year contracts. Average contract duration extended to 22 months as of Q3 FY2025, up from approximately 18 months in FY2024 [S3].

**Land-and-expand mechanics:** [FACT] Amplitude's go-to-market is structured as a land-and-expand motion. Customers typically begin with core product analytics (often via a free tier or small initial contract) and expand over subsequent contract cycles by adding Session Replay, Feature Experimentation, Audience Activation, Guides & Surveys, or the Amplitude CDP. As of Q3 FY2025, 70% of ARR came from customers using multiple products, up from an estimated 40–50% two years prior [S3].

**Dollar-Based Net Revenue Retention (NRR):** [FACT] NRR measures the expansion (upsell + cross-sell) minus contraction and churn from the existing customer base over a trailing 12-month period. Amplitude's NRR hit a trough of 96% in Q2 FY2024 — reflecting the 2022–2024 enterprise software spending retrenchment — and recovered to 104% by Q2 FY2025 and 104% for FY2025 full-year [S1, S3]. An NRR above 100% means existing customers are growing their spend faster than they churn, generating organic revenue growth from the installed base alone.

**Remaining Performance Obligations (RPO):** [FACT] RPO reached $391.9 million as of Q3 FY2025, up 37% year-over-year — growing materially faster than ARR growth of 16%. This indicates Amplitude is signing longer-duration, larger-value contracts at an accelerating pace, providing forward revenue visibility [S3].

**Revenue recognition:** Amplitude recognizes subscription revenue ratably over the contract period. This means a large enterprise deal signed in Q4 FY2025 contributes relatively little to FY2025 revenue but will be fully recognized in subsequent periods — creating a lag between bookings momentum and reported revenue [S1].

---

#### 4. Customer Segmentation

**By spend tier:**

- [FACT] **Enterprise (≥$100K ARR):** 698 customers as of December 31, 2025 — up 18% year-over-year from 591. This cohort generates 78% of total ARR ($285M+ of $366M) and is the primary driver of land-and-expand economics [S1].
- [FACT] **High-value enterprise (≥$1M ARR):** 56 customers as of FY2025, up 33% from 42 in FY2024. These megadeals signal Amplitude's ability to win significant platform-level commitments from large organizations [S1].
- [FACT] **SMB/self-serve (<$100K ARR):** Remaining ~4,099 customers (4,797 total minus 698 enterprise). These customers are important for top-of-funnel volume but contribute only ~22% of ARR at the blended average.
- [EST] Enterprise accounts ($100K+ ARR) carry an average ACV of approximately $408K ($285M / 698 customers), while the SMB cohort averages approximately $20K per customer — a 20x ACV gap that motivates Amplitude's upmarket push.

**By vertical:** [FACT] Amplitude's disclosed customer logos span B2B SaaS, fintech/payments, consumer media/streaming, e-commerce, and healthcare (HIPAA-compliant). Notable named customers include Atlassian, NBCUniversal, Under Armour, and Shopify. The company claims coverage of 27 Fortune 100 companies [S1, S3].

**By geography:** [FACT] United States generated $209.8 million (61%) and International generated $133.4 million (39%) of FY2025 revenue. US revenue grew 17% year-over-year; International grew 12%, suggesting the US enterprise segment is expanding faster than international [S1].

**By product maturity cycle:** [JUDGMENT] The customer base is bifurcating into (a) legacy single-product analytics accounts that need to be migrated onto multi-product bundles to improve NRR, and (b) new multi-product platform accounts with structurally higher retention. Amplitude's ability to upgrade the former cohort is the central operational challenge of the next 12–24 months.

---

#### 5. Go-to-Market

**Product-Led Growth (PLG):** [FACT] Amplitude maintains a free tier (up to a threshold of events/MTUs) that allows product managers, data analysts, and engineers to begin using the platform without budget approval. This drives organic top-of-funnel awareness within prospective customers. The PLG motion is self-serve and does not require direct sales involvement [S1, S3].

**Enterprise sales overlay:** [FACT] Above the free tier, Amplitude deploys a direct sales force with dedicated account executives for enterprise accounts. Sales and marketing expenses were $188.0 million in FY2025 (54.8% of revenue) — elevated relative to Amplitude's growth rate, reflecting ongoing investment in enterprise sales capacity [S1]. The company targets VP of Product, CPO, CTO, and Chief Data Officer buyers.

**Solution partners and technology integrations:** [FACT] Amplitude maintains a partner ecosystem of systems integrators, technology partners (Snowflake, Databricks, Salesforce), and analytics consulting firms that assist with implementation and ongoing use. Partner-sourced deals reduce the direct sales cost-per-customer for complex enterprise deployments [S1].

**Outcome-based selling:** [FACT] A commissioned Forrester Total Economic Impact study cited by management claims a 6-month payback period for Amplitude platform customers. This ROI framework is used as a sales tool to justify premium pricing vs. lower-cost alternatives [S3].

**International GTM:** [FACT] Amplitude operates seven global offices as of FY2025. International revenue grew 12% in FY2025, slower than US growth of 17%, suggesting current GTM investment is concentrated domestically. No specific international ARR or customer count breakdowns are disclosed by management [S1].

---

#### 6. Platform Expansion ("Year of the Platform")

In FY2025, CEO Spenser Skates branded the company's strategy as the "Year of the Platform" — explicitly positioning Amplitude as a unified analytics platform rather than a single-use product analytics tool. The strategic thesis is that customers will consolidate 4–6 disconnected point solutions (analytics, session replay, A/B testing, feature flags, surveys, CDP) under one Amplitude instance, generating higher ACV, longer contract terms, and better NRR [S3].

**Product inventory as of FY2025:**

| Product | Launch Status | Competitive Target | Multi-Product Contribution |
|---|---|---|---|
| Analytics (Core) | Mature flagship | Mixpanel, Heap, PostHog | Baseline — all customers |
| Session Replay | Launched Feb 2024 | FullStory, LogRocket, Hotjar | Growing adoption; bundled with analytics |
| Feature Experimentation | Existing; scaled | Optimizely, Statsig, LaunchDarkly | Established; engineering buyer |
| Web Experimentation | Existing | Optimizely | Smaller; complements Feature Exp. |
| Amplitude CDP | Existing; expanding | Segment (Twilio), mParticle | Upstream data ownership story |
| Audience Activation | Existing | Adobe Audience Manager | Tied to CDP; downstream activation |
| Guides & Surveys | Launched 2024 | Pendo, Appcues | Fastest adoption per Q1 FY2025 commentary |
| AI Agents | Launched 2025 | Emerging — no clear incumbent | Pre-revenue; adoption phase |
| Amplitude MCP | Launched 2025 | N/A (enabling layer) | LLM integration layer; strategic |
| AI Visibility / AI Feedback | Launched 2025 | Emerging AI monitoring tools | Early stage |
| Automated Insights | Existing; enhanced | BI tool AI features | Bundled; not separately priced |

[FACT, S1, S3]

**Multi-product ARR evidence:** [FACT] 70% of ARR came from customers using multiple Amplitude products as of Q3 FY2025. Management frames this as the primary evidence that the platform strategy is working. Multi-product customers have materially higher NRR (estimated >110%) vs. single-product customers, though Amplitude does not disclose this split explicitly [S3].

**AI acquisitions:** [FACT] In 2025, Amplitude completed three acquisitions to accelerate AI capabilities: Command AI (in-product AI assistant), Craffle (AI analytics enhancement), and Anari (AI-powered analytics). These acquisitions contributed to the $10.3M increase in R&D personnel costs in FY2025, partially offset by an $11.8M decrease in SBC expense (prior-year SBC had been elevated due to Command AI acquisition-related acceleration) [S1].

**MCP integration:** [FACT] Amplitude launched the Amplitude MCP (Model Context Protocol) layer in 2025, designed to expose Amplitude behavioral data to external AI agents and large language models. This positions Amplitude as a data source for AI-powered workflows, extending the platform's relevance into the emerging agentic AI ecosystem [S1].

[JUDGMENT] The platform expansion is the correct strategic direction but carries execution risk: each new product adds implementation complexity for customers, competes with established best-of-breed vendors in its respective category, and may dilute the brand clarity that Amplitude has built in product analytics. The key test over the next 18–24 months is whether multi-product attachment rates translate into sustainably higher NRR (>105%) or whether platform breadth creates implementation fatigue and feature sprawl churn.

---

#### 7. Key Dependencies and Risks to Model

**Third-party cloud infrastructure:** [FACT] Amplitude's cost of revenue includes material payments to cloud providers (primarily AWS and GCP) for compute, storage, and data transfer. In FY2025, third-party hosting costs increased $6.0 million year-over-year, the largest single driver of cost-of-revenue growth [S1]. As Session Replay and CDP storage requirements grow, hosting costs could compress gross margins below the current 74% GAAP / 76% non-GAAP level.

**Data pipeline and integration dependencies:** [FACT] Amplitude's warehouse-native integrations depend on API access and partnership arrangements with Snowflake, Databricks, and Google BigQuery. Changes in these platform vendors' pricing or data-sharing policies could affect the competitiveness of Amplitude's warehouse-native offering [S1, S2].

**AI commoditization risk:** [JUDGMENT] The most structurally significant long-term risk is the potential for warehouse-native AI query tools (Snowflake Cortex, BigQuery ML, Databricks AI) to replicate enough of Amplitude's analytics GUI functionality that buyers question the value of a standalone analytics layer. This substitution risk is a 3–5 year horizon concern rather than immediate, but it shapes how Amplitude must position its AI-native differentiation [S2].

**Free-tier competition:** [FACT] Google Analytics 4 + Firebase Analytics offer behavioral tracking at $0 for web and mobile, anchoring price expectations for the SMB and startup segment. PostHog provides an open-source self-hosted alternative with capabilities that now overlap substantially with Amplitude's core product, and has grown to deployments on 5,330+ of the top 1 million websites [S2].

**Stock-based compensation (SBC) dilution:** [FACT] SBC expense was $92.1 million in FY2025 — equal to 26.8% of total revenue and 36.3% of gross profit. Weighted average diluted shares outstanding grew from 111.4 million in FY2022 to 132.0 million in FY2025, representing 18.5% dilution over three years. At current share repurchase pace ($50M authorized in Q1 FY2025), buybacks are insufficient to offset the pace of new share issuance [S4].

**Dual-class share structure:** [FACT] AMPL went public via direct listing in September 2021 with a dual-class share structure, concentrating voting control with founders. This reduces minority shareholder influence over capital allocation, M&A, and executive compensation decisions [S1].

**NRR recovery fragility:** [JUDGMENT] The NRR recovery from 96% (Q2 FY2024) to 104% (FY2025) is encouraging but occurred in a relatively benign macro environment. Enterprise software budgets remain discretionary, and a re-tightening of IT spending — particularly for analytics tools with easily deferrable renewal dates — could reverse the expansion trend. The 2022–2024 NRR trough demonstrated that Amplitude's customer base is not immune to budget retrenchment cycles.

**Customer concentration (low but worth noting):** [FACT] No single customer accounted for more than 10% of FY2025 revenue [S1]. This is a positive structural attribute — AMPL is not dependent on any single enterprise relationship for a material portion of revenue.

---

#### 8. Source Index

| Source ID | Description |
|---|---|
| [S1] | Amplitude, Inc. FY2025 10-K Annual Report (filed February 19, 2026; period ending December 31, 2025; accession 0001193125-26-057847). Full filing at https://www.sec.gov/Archives/edgar/data/1866692/000119312526057847/ampl-20251231.htm |
| [S2] | AMPL Competitive Landscape research file (June 2026); data from Mixpanel, Heap/Contentsquare, PostHog, Google Analytics 4, Adobe Analytics comparative analysis |
| [S3] | AMPL Investor Presentation Summary (compiled from Q4 FY2024, Q1–Q3 FY2025 earnings releases and calls; June 2026) |
| [S4] | AMPL XBRL Data Summary (SEC EDGAR XBRL API, CIK 0001866692; retrieved June 14, 2026) |
| [S5] | AMPL Market Overview: Product Analytics & Digital Experience Intelligence (June 2026) |

---

*Step 01 prepared under coverage-next-full protocol. No earnings transcripts used. Data sources: SEC filings, XBRL, investor materials, industry research.*

## Recent Catalysts

---
source: coverage-next-full
ticker: AMPL
company: Amplitude, Inc.
step: 12
title: Bull/Bear Catalysts
date: 2026-06-14
---

### Step 12 — Bull/Bear Catalysts: Amplitude, Inc. (AMPL)

*Coverage-next-full path: filings + consensus + industry research only. No earnings transcript analysis performed. Bull/bear debate is inferred from SEC filings (10-K FY2025, 8-K Q1 FY2026), analyst consensus data, press releases, and industry research. Transcript-derived color on management tone and Q&A dynamics is not included.*

---

#### 1. Market Debate Overview

Amplitude trades at $6.88 — approximately 85% below its IPO-high near $50 and down 40% year-to-date as of June 2026 [S3]. The stock is not merely a "de-rated growth stock" — it is the focus of an active securities fraud investigation and carries a market capitalization of $912M against $374M in ARR, implying an EV/ARR multiple of approximately 2.0x. For context, the median healthy growth SaaS company traded at 8–15x ARR multiples at the peak of 2021; a 2x multiple historically prices in material fundamental risk or near-terminal growth expectations.

The central market disagreement is not about whether Amplitude has a viable product — it manifestly does, with 27 Fortune 100 customers, 698 accounts above $100K ARR, and NDRR recovering to 106% as of Q1 FY2026 [S1, S3]. The disagreement is about whether Amplitude is a durable standalone platform business with pricing power, or a commoditizing analytics tool that will be disintermediated by AI-native data platforms within three to five years.

**What the bull believes that the bear denies:** Platform consolidation is a real, durable trend — customers are actively replacing 4–6 point solutions with Amplitude's integrated stack, evidenced by multi-product customers growing from 10% to 24% of ARR in a single year [S3]. At 2x EV/ARR with improving NRR and three consecutive years of positive FCF, the stock is priced for failure at a company that is generating cash and growing 17% ARR YoY.

**What the bear believes that the bull denies:** The May 2026 guidance cut is not a one-time calibration — it exposes structural cost uncertainty as AI becomes the company's primary product surface. SBC at $92M on $343M revenue means real shareholders fund employee compensation through dilution while GAAP losses persist. And the category is ultimately too small and too easily replicated by Google, Snowflake, or open-source alternatives to justify a standalone premium valuation [S3, S4].

---

#### 2. Bull Case Narrative

**NRR Recovery Is Structural, Not Cyclical.** Amplitude's NRR fell to 96% at trough in mid-2024, triggering investor concern about the durability of the subscription model [S2]. The recovery to 101% in Q1 2025 and 104% in Q4 2025 year-end, reaching 106% by Q1 FY2026, is not simply a macro tailwind — it reflects a deliberate platform consolidation strategy [S1, S3]. As of Q1 FY2026, customers using five or more products now represent 24% of ARR, up from 10% in Q1 2025. Multi-product customers have structurally higher NRR because each additional module (Session Replay, Feature Experimentation, AI Agents) creates both revenue expansion and switching-cost deepening. The bull thesis holds that NRR at 106% is the floor of a trend that can reach 110%+ as the multi-product attach rate continues rising.

**Platform Expansion Is Real and Accelerating.** Amplitude's product surface now spans eleven distinct capabilities: core analytics, Session Replay (launched February 2024), Feature Experimentation, Web Experimentation, Audience Activation, Guides & Surveys, AI Agents, Amplitude MCP, AI Visibility, AI Feedback, and Automated Insights [S1]. The "Year of the Platform" strategy (2025) is producing measurable results: 70% of ARR now comes from multi-product customers, and customers with ARR above $1M grew 33% YoY to 56 accounts [S1]. Each new product module expands the total available wallet at each customer account, improving both NRR expansion potential and churn resistance.

**Warehouse-Native Architecture Is the Correct Strategic Bet for the AI Era.** The most credible disruption scenario for Amplitude involves AI-powered natural language querying of behavioral data already in data warehouses — which would eliminate the need for a dedicated analytics GUI. Amplitude's response — building native integrations with Snowflake, BigQuery, and Databricks, and positioning itself as the behavioral analytics layer that sits atop warehouse infrastructure — converts this threat into a distribution advantage. Customers who run behavioral data in Amplitude can access it from the warehouse; customers whose data is in the warehouse can sync it into Amplitude. The bull reads warehouse-native positioning as Amplitude becoming *essential* plumbing in the AI-era data stack, not a casualty of it [S2, S4].

**Non-GAAP Profitability Achieved and FCF Structural.** FY2025 marked Amplitude's first full year of non-GAAP operating profitability ($1.2M) [S1]. More importantly, free cash flow was $23.5M in FY2025 — the company's second consecutive year of positive FCF and a record for the company. FCF has been positive for three consecutive fiscal years (FY2023, FY2024, FY2025), demonstrating that the subscription business generates real cash without requiring additional equity capital. At $252.5M in total liquid assets against minimal debt, Amplitude has a multi-year cash runway with no near-term financing risk [S1].

**Valuation Is Near Trough Multiples for a Growing SaaS Company.** At 2.1x EV/Revenue (TTM) and approximately 2.0x EV/ARR, Amplitude trades at multiples typically associated with companies facing terminal decline or structural disruption. ARR is growing at 17% YoY ($374M as of Q1 FY2026) with improving NRR and a $427M remaining performance obligation (+31% YoY), providing visibility into near-term revenue recognition [S3]. For a company with 15–17% topline growth, positive FCF, improving NRR, and a No. 1 G2 position in its core category, a 2x ARR multiple implies a market expectation of deceleration that the underlying data does not yet support [S2, S3].

**Insider Conviction Visible in Open-Market Buys.** An Amplitude CFO open-market stock purchase of approximately $300K at prices meaningfully above trough levels signals that insiders — who have access to non-public operational data — view the current share price as undervalued. Open-market buys (as opposed to option exercises or scheduled RSU vesting) are the highest-conviction insider signal available [S3].

---

#### 3. Bear Case Narrative

**AI Commoditization Will Erode the Core Value Proposition.** The most powerful structural bear argument is that product analytics — answering "who did what, when, and in what sequence?" — is becoming a commodity capability that LLMs and data warehouse query interfaces can increasingly replicate at near-zero marginal cost. Google (BigQuery ML + GA4), Snowflake (Cortex AI), Databricks (Delta Live Tables + AI assistants), and Metabase (AI-powered natural language querying) are all building toward the same destination: a product manager asking a warehouse-native AI "show me 30-day retention for users who completed onboarding in the last 90 days" and receiving a formatted answer without opening Amplitude [S2, S4]. Amplitude's Behavioral Graph provides meaningful depth advantages today in complex multi-event behavioral sequencing — but that moat narrows as AI systems improve at structuring warehouse queries and interpreting user journey data. On a five-year view, the category may split into: (1) a free/embedded capability within data platforms for standard use cases, and (2) a niche deep-analysis layer for a small subset of data-mature organizations. Amplitude's $400M ARR target assumes it serves a much broader market than scenario (2) accommodates.

**SBC Dilution Is Predatory and Ongoing.** Amplitude spent $92M in stock-based compensation in FY2025 against $343M in total revenue — a 26.8% SBC-to-revenue ratio [S1]. This is not a growth-phase anomaly: the company has maintained elevated SBC throughout its history and the diluted share count has grown approximately 19% since IPO. Non-GAAP earnings ($0.04 per diluted share for FY2025) exclude SBC entirely, presenting a flattering picture of economics that does not reflect the real cost of retaining the talent that generates those economics. Real shareholders — those who bought stock at IPO or in the secondary market — are effectively funding employee compensation through the dilution of their ownership, not through earnings. The bear argues that until GAAP profitability is achieved and SBC is brought below 15% of revenue, the stock represents a wealth transfer mechanism from public shareholders to employees, not a genuine equity investment. FY2026 guidance implies SBC declines to approximately $85M — still above 21% of projected revenue at the guidance midpoint of $400M [S3].

**May 2026 Guidance Cut Reveals Forecasting and Cost Structure Uncertainty.** When Amplitude cut its FY2026 non-GAAP operating income guidance from $7M–$13M to $2.5M–$6.5M on May 6, 2026 — a reduction of $5–6M at midpoint — the stock fell 21% in a single session [S3]. The stated cause was higher-than-modeled AI inference costs from the Statsig partnership and AI Agents deployment. The bear reads this not as a one-time calibration, but as evidence that management cannot model the cost structure of its own AI product suite with sufficient precision to provide reliable guidance. If AI becomes 30–40% of Amplitude's product offering over the next 18 months — as the company's product roadmap suggests — and inference costs are structurally unpredictable, the non-GAAP profitability framework that underpins the bull thesis is built on uncertain foundations. The Pomerantz LLP investigation adds a second layer: it implies that investors may have been given misleading forward projections prior to the guidance cut, creating litigation overhead and management distraction at a critical execution phase [S3].

**NRR at 106% Reflects a Below-Average Expansion Machine.** Best-in-class enterprise SaaS companies with genuine land-and-expand mechanics — Snowflake, Datadog, CrowdStrike at their growth peaks — maintained NRR in the 120–130%+ range, meaning the installed base grew 20–30% annually from existing customers alone. Amplitude's 106% NRR, while recovered from its 96% trough, still implies only 6 cents of expansion per dollar of existing ARR, which is insufficient to drive high topline growth from installed-base dynamics alone [S3]. The company remains dependent on new logo additions to sustain 15–17% revenue growth — and new logo growth in the current macro environment is harder to achieve than expansion from embedded customers. The bear argues that without NRR reaching 110%+, Amplitude is running an inherently fragile growth model that requires continuous new customer acquisition against intensifying competition.

**Competitive Intensity Will Intensify as AI Reduces Development Cost Barriers.** PostHog is already deployed on more sites than Adobe Analytics and provides near-full product analytics parity at zero license cost for self-hosted deployments [S4]. As LLM-assisted development reduces the engineering cost of building analytics interfaces, the barriers to building competitive product analytics features continue to fall. This is not a theoretical risk: the product analytics market has twelve or more credible vendors competing across overlapping use cases today, and new entrants with AI-accelerated product development will emerge from the current wave of foundation model infrastructure investment [S2, S4].

---

#### 4. Key Debates

**Debate 1: Is NRR 106% the New Floor or a Cyclical Bounce?**

The bull views 106% NRR as the structural floor of an improving trend anchored in multi-product adoption — as customers add Session Replay, Feature Flags, and AI Agents modules, NRR mechanically improves and churn risk falls. The 5+ product cohort growing from 10% to 24% of ARR in a single year provides direct evidence of this dynamic [S3]. The bear counters that 106% NRR reflects the current macro environment (tech spending recovery in 2025–2026) rather than structural stickiness — and that the next economic contraction will re-expose the discretionary nature of product analytics spend, reproducing the 2024 NRR trough. The resolution likely depends on whether multi-product attach rates continue compounding (bull) or plateau as customers resist additional modules during budget cycles (bear).

**Debate 2: Does Warehouse-Native Architecture Defend Against or Accelerate Disintermediation?**

The bull argues warehouse-native architecture makes Amplitude indispensable: by sitting cleanly atop Snowflake/BigQuery/Databricks, Amplitude becomes the behavioral analytics query layer for the AI-era data stack — customers need a purpose-built interface for behavioral analysis even when raw data lives in the warehouse [S4]. The bear argues the opposite: warehouse-native architecture accelerates disintermediation by commoditizing the data transport problem (data is already in the warehouse), leaving only the analytics GUI layer as Amplitude's differentiator — and that GUI layer is precisely what AI-native natural language interfaces can replace. The outcome hinges on whether behavioral sequence analysis (multi-step user journeys, cohort behavioral attribution) remains complex enough to require a purpose-built tool indefinitely, or whether improving AI query sophistication eliminates that requirement.

**Debate 3: Is SBC Declining Meaningfully or Remaining Structurally Elevated?**

The bull observes that SBC declined from a peak level (elevated by one-time Command AI acquisition vesting) and guidance implies further reduction toward approximately $85M in FY2026 — a demonstrable improvement trend. As the company scales revenue, SBC as a percentage of revenue will fall even if the absolute dollar amount declines modestly. The bear notes that $85M SBC on $400M projected revenue is still 21%+ of revenue — far above the 10–12% SBC-to-revenue levels that characterize sustainably profitable SaaS companies — and that the gap from GAAP to non-GAAP ($88.5M GAAP net loss vs. near-breakeven non-GAAP in FY2025) is almost entirely explained by SBC rather than one-time items [S1]. Share count dilution since IPO (~19%) has transferred substantial value from public equity holders to employees.

**Debate 4: Is the May 2026 Guidance Cut a One-Time AI Cost Calibration or Structural Margin Constraint?**

The bull accepts management's framing: AI inference costs were undermodeled in the initial FY2026 guidance, the Statsig integration brought unexpected compute requirements, and now that those costs are visible, guidance has been recalibrated with appropriate buffers. Non-GAAP operating profitability is maintained (guidance: $2.5M–$6.5M), and FCF will recover in back-half-weighted seasonality [S3]. The bear challenges whether any guidance for AI inference costs can be reliable when token consumption and model serving costs are functions of both adoption growth and underlying model pricing — factors outside Amplitude's control. If AI Agents becomes Amplitude's primary growth vector (as the product roadmap suggests), the cost basis of the most rapidly growing revenue segment is structurally unpredictable. The BofA downgrade captured this concern explicitly, cutting its price target from $10 to $8 on execution risk grounds [S3].

---

#### 5. Catalyst Inventory (12–18 Month Horizon)

**Positive Catalysts:**

- *NRR reaching 108–110%:* Would demonstrate that multi-product NRR improvement is durable through a full business cycle, validating the platform consolidation thesis. Market would likely re-rate EV/ARR toward 3–4x.
- *Statsig partnership delivering $16M ARR:* If the partnership closes as projected in FY2026, it validates the channel economics and provides a proof-of-concept for similar experimentation platform partnerships.
- *AI Agents achieving $20M+ ARR:* An identifiable, material revenue contribution from AI products would validate the AI growth vector and justify the inference cost investment thesis.
- *Q2 or Q3 FY2026 revenue beats with raised full-year guidance:* Consecutive beats would rebuild investor confidence in management's forecasting precision following the May 2026 guidance cut.
- *Pomerantz investigation formally dropped or not escalated to class action:* Removes the litigation discount embedded in current valuation and eliminates management distraction risk.
- *M&A at a premium:* At $912M market cap and $252.5M in cash, Amplitude is technically acquirable by a larger platform vendor (Salesforce, Adobe, Databricks) at a meaningful premium to current price, though no acquisition speculation is currently evident in analyst coverage [S3].

**Negative Catalysts:**

- *NRR reversal below 100%:* Would signal that the multi-product platform thesis is not protecting existing ARR cohorts and would likely trigger a significant re-rating lower.
- *AI inference cost escalation beyond FY2026 guidance:* If Q2/Q3 FY2026 non-GAAP operating income misses guidance midpoint due to AI costs, investor confidence in the profitability timeline collapses.
- *Class action certification in Pomerantz investigation:* Shifts from tail risk to active liability, with settlement costs, defense costs, and management distraction compressing near-term execution.
- *Enterprise churn from AI tool consolidation:* If 5–10 of the 698 $100K+ ARR accounts consolidate analytics into their data platform or warehouse-native tools, investor scrutiny of the cohort churn rate intensifies.
- *CEO or CTO departure:* Co-founder departures in founder-led SaaS companies are disproportionately disruptive and typically trigger stock declines of 15–30%+ at the announcement.
- *PostHog enterprise tier launch gaining material traction:* If PostHog achieves credible enterprise certifications and begins winning in Amplitude's $100K+ ARR segment, it would signal the commoditization thesis is advancing ahead of schedule [S4].

---

#### 6. Bull Case — Summary

**Bull Case:**
- NRR is structurally improving on the back of multi-product adoption (5+ product customers grew from 10% to 24% of ARR YoY), creating a compound upsell flywheel that is increasingly difficult to reverse as switching costs deepen with each additional module deployed.
- At 2.1x EV/Revenue with 17% ARR growth, three consecutive years of positive FCF, and a $427M RPO (+31% YoY), Amplitude is priced for failure at a business demonstrably generating cash and booking multi-year enterprise commitments — creating a substantial margin of safety relative to fundamental trajectory.
- Warehouse-native architecture and AI Agents (via MCP integration) position Amplitude as essential behavioral data infrastructure for the AI era rather than a legacy GUI at risk of disintermediation, with the platform breadth (11 modules) creating a compelling consolidation narrative that no pure-play competitor can match.

---

#### 7. Bear Case — Summary

**Bear Case:**
- SBC of $92M on $343M revenue ($85M+ guided for FY2026) means GAAP losses will persist for the foreseeable future, and real shareholders are funding employee compensation through ~19% post-IPO dilution rather than earning a return on equity — the non-GAAP profitability framing obscures this wealth transfer.
- The May 2026 guidance cut demonstrated that management cannot forecast the cost structure of its own core growth product (AI Agents/inference costs), and the Pomerantz fraud investigation implies investors may have been given misleading forward projections — creating litigation liability, management distraction, and a credibility discount that will take multiple clean quarters to rebuild.
- AI-native querying in Snowflake, BigQuery, and emerging open-source tools (PostHog's Max AI) is advancing toward the point where product managers can answer behavioral analytics questions directly from warehouse data, threatening to compress Amplitude's addressable market to a smaller subset of data-mature users willing to pay a premium for depth — not the 4,797-customer broad platform Amplitude is pricing itself to be.

---

#### 8. Source Index

| Code | Source |
|------|--------|
| [S1] | Amplitude, Inc. Form 10-K for Fiscal Year Ended December 31, 2025 (filed February 19, 2026). Accession No. 0001193125-26-057847. SEC EDGAR. |
| [S2] | Product Analytics & Digital Experience Intelligence — Market Overview. Internal research compilation. June 2026. Citing: Grand View Research; Mordor Intelligence; The Business Research Company; Quantum Metric; SecurePrivacy.ai. |
| [S3] | Amplitude, Inc. — Market Data & Analyst Consensus. Internal compilation. June 14, 2026. Sources: Yahoo Finance; StockAnalysis.com; Pomerantz LLP Press Release (PR Newswire, May 2026); BofA Downgrade (Proactive Investors, May 2026); SEC EDGAR 8-K Q1 FY2026 (Accession 000119312526209264); SEC EDGAR 8-K Q4 FY2025 (Accession 000119312526057052). |
| [S4] | Amplitude (AMPL) — Competitive Landscape. Internal research compilation. June 2026. Citing: G2 Summer 2025 Report; Forrester Wave Digital Analytics 2025; PostHog deployment data (October 2024); aakashg.com product analytics market deep dive; adapty.io; posthog.com/compare. |

## Full Investment Thesis (Premium)

The full research tier adds these thesis-critical dimensions:

- Moat Analysis — durable competitive advantages, switching costs, network effects
- Investment Thesis — variant perception, what has to be true, why market may be wrong
- Bull / Base / Bear Scenarios — probability weights, catalysts, price targets
- Risk Register — macro, competitive, execution, regulatory risks with materiality ratings
- Management Quality — capital allocation track record, incentive alignment
- DCF Valuation — 10-year model with sensitivity matrix

**API endpoint:** GET /api/v1/research/AMPL/memo

## Navigation

- Overview: /stocks/ampl
- Financials: /stocks/ampl/financials
- Thesis (this page): /stocks/ampl/thesis
- Investment Memo: /memo/ampl
- Coverage universe: /stocks