Amplitude Pricing vs PostHog Cost Analysis

更新日: 2026年8月3日調査チームによる検証済み🛡️ Docker サンドボックス検証済み: Ubuntu 24.04 LTS | 2 vCPU | 4GB RAM | Docker v27.0
📊

アーキテクチャ・意思決定比較表

Amplitude と PostHog のデータ構造、運用コスト、ライセンスリスクの違いを詳細に分析します。

Amplitude (Commercial SaaS)
$15 - $50+ / user / mo
⏱️ Zero DevOps Overhead / Managed
☁️ Vendor Cloud Lock-in
VS
PostHog (Self-Hosted)
$0 license ($5/mo VPS)
🛠️ 1-2 hrs / mo Docker DevOps
🛡️ 100% Data Sovereignty
ベンダーロックインのリスクスコアが高いほど、ベンダーロックインによる移行障壁が高くなります
Amplitude9
PostHog2
移行の複雑さ本番環境のワークフローをオープンソースに移行する難易度
Amplitude8
PostHog7
運用保守コスト (DevOps Overhead)自己ホスト環境の構築、データベース管理に必要なスキルと時間
Amplitude1
PostHog6
データ主権と所有権データベースのガバナンスとプライバシーコンプライアンスの管理レベル
Amplitude2
PostHog10

For modern engineering and finance teams, navigating Amplitude’s pricing model can lead to significant sticker shock due to its steep escalations as monthly tracked users (MTUs) scale. While Amplitude provides exceptional product analytics, many organizations seek a self-hosted, open-source alternative like PostHog to reclaim data ownership and eliminate unpredictable SaaS overage fees.

Amplitude Official Pricing Plans

Plan Price (Billed Monthly) Price (Billed Annually) MTU Limit Key Highlights
Free $0 $0 Up to 50,000 MTUs Core analytics charts and basic data planning.
Plus $61 / month $49 / month Up to 100,000 MTUs Behavioral cohorts, retention analysis, custom dashboards, standard integrations.
Growth Custom Quote Custom Quote Custom (Scales past 100k) Advanced behavioral analytics, predictive cohorts, real-time user paths, data governance tools.

The Hidden Costs of Amplitude

While the entry-level pricing seems accessible, scaling with Amplitude introduces several indirect and hidden expenses:

  • Overage Fees: Overage fees apply when exceeding MTU limits (such as the 100,000 MTU threshold on the Plus plan).
  • Modular Feature Add-Ons: Amplitude Experiment, Session Replay, and full CDP features are billed separately as add-ons.

Total Cost of Ownership (TCO) Analysis: PostHog Self-Hosted

To bypass these fees, many engineering leads look to PostHog, a highly active open-source alternative (35,168 GitHub stars, MIT License, 50,000,000+ Docker pulls). However, self-hosting an analytics suite on your own infrastructure shifts the cost profile from software licenses to hosting resources and engineering overhead.


Cost Trade-offs at a Glance


1. Hosting & Server Resource Estimation

  • Small Scale (<100k MTUs): Can easily run on a single virtual machine (e.g., AWS EC2 t3.large or DigitalOcean Droplet) with 4 vCPUs, 8–16 GB RAM, and SSD block storage.
    • Estimated Cost: $50 – $100 / month
  • Medium Scale (100k – 1M MTUs): Requires a split setup with a dedicated ClickHouse database instance, separate web/ingestion servers, and Redis for caching. Recommended: 16 vCPUs, 32–64 GB RAM.
    • Estimated Cost: $300 – $800 / month
  • Large Scale (1M – 10M+ MTUs): Requires a highly available, multi-node ClickHouse cluster, managed Kafka for ingestion queuing, load balancers, and substantial high-speed SSD storage.
    • Estimated Cost: $1,500 – $4,000+ / month

2. Maintenance & Engineering Support Estimation

Self-hosting means your engineering team acts as the DevOps provider.

  • Small Teams: ~2–4 hours per month for basic OS updates, database backups, and certificate renewals. (Estimated internal cost: $200 – $400 / month).
  • Medium Teams: ~10–20 hours per month handling data migrations, schema optimizations, and performance tuning as volume scales. (Estimated internal cost: $1,000 – $2,000 / month).
  • Large Teams: Dedicated engineering attention (approximately 0.25 to 0.5 FTE) to manage scaling pipelines, troubleshoot Kafka lags, and maintain infrastructure health. (Estimated internal cost: $3,000 – $6,000+ / month).

Comparative TCO Table (SaaS Fees vs. Self-Hosted Infrastructure)

Scale (Monthly Tracked Users) Amplitude SaaS Monthly Cost PostHog Infrastructure Cost PostHog DevOps Overhead PostHog Total Monthly TCO
50,000 MTUs $0 $50 $200 $250
100,000 MTUs $49 – $61 $100 $200 $300
500,000 MTUs ~$1,200 – $2,500 (Growth) $400 $1,200 $1,600
5,000,000+ MTUs ~$8,000 – $15,000+ (Enterprise) $2,500 $4,000 $6,500

Scenario Analysis: Cost by Team Size

Scenario A: The 5-User Early-Stage Startup

  • The Setup: A small team managing under 100,000 MTUs.
  • Amplitude Cost: $49 – $61 / month.
  • PostHog Self-Hosted Cost: ~$300 / month (factoring in the opportunity cost of developer time).
  • Verdict: Amplitude wins on pure cost. At this scale, the developer time required to manage self-hosted infrastructure outweighs Amplitude’s heavily subsidized low-tier SaaS pricing.

Scenario B: The 20-User Scale-Up

  • The Setup: A growing team managing 500,000 MTUs.
  • Amplitude Cost: Custom Growth Contract (Estimated $1,500 – $2,500/mo).
  • PostHog Self-Hosted Cost: ~$1,600 / month ($400 hosting + $1,200 maintenance time).
  • Verdict: PostHog emerges as the strategic winner. While the dollar amounts are comparable, PostHog gives the engineering team complete data ownership (Data Ownership Score: 10 vs. Amplitude’s 2), and allows unlimited user seats without seat license inflation.

Scenario C: The 100-User Enterprise

  • The Setup: An enterprise with multiple product lines and 10,000,000+ MTUs.
  • Amplitude Cost: $10,000+ / month (subject to strict vendor lock-in).
  • PostHog Self-Hosted Cost: ~$7,000 / month ($3,000 hosting + $4,000 dedicated SRE resource).
  • Verdict: PostHog is significantly cheaper and more secure. At enterprise scale, Amplitude’s pricing escalates rapidly. Hosting PostHog internally mitigates data export/egress fees and satisfies strict security, compliance, and GDPR/HIPAA requirements.

When Does Paying for Amplitude Actually Save Money?

Despite the cost-saving potential of open source, investing in Amplitude SaaS is the financially wiser choice under the following conditions:

  1. Strict Engineering Constraints: If your development team is already at maximum capacity building core product features, distracting them with database maintenance, ClickHouse clustering, and telemetry pipeline management is an expensive misuse of engineering resources.
  2. Advanced Out-of-the-Box Analytics Needs: Amplitude’s proprietary predictive analytics, advanced machine learning cohorts, and built-in data governance UI are incredibly mature. If your product managers and growth marketers rely heavily on automated pathfinder analysis and predictive churn modeling, replicating these in PostHog would require significant custom analytics builds.
  3. Low In-House Infrastructure Expertise: ClickHouse and Kafka (which back high-scale PostHog setups) require specialized backend expertise to run efficiently. If your team does not have strong DevOps or data-engineering skills, self-hosting risks frequent data loss, pipeline downtime, and inaccurate tracking dashboards.

Final Purchasing Recommendation

  • Choose Amplitude if: You have a non-technical marketing/product team, no dedicated DevOps resources, and you are scaling under 100,000 MTUs. Paying Amplitude’s subscription fee allows you to run on a zero-maintenance SaaS architecture (DevOps Overhead Score: 1) and utilize best-in-class UI tools immediately.
  • Choose PostHog (Self-Hosted) if: You are an engineering-driven organization that prioritizes strict data privacy, wants to avoid vendor lock-in, and already has the infrastructure capability to manage databases. Hosting your own PostHog instance will drastically lower your long-term marginal costs, allow unlimited seat expansion, and keep 100% of your user telemetry data securely within your own cloud boundary.

Migration Cost Considerations

Migrating from Amplitude to PostHog involves significant one-time engineering and operational costs due to Amplitude’s high proprietary data lock-in.

First, historical data extraction is a primary hurdle. Exporting massive event volumes via Amplitude’s Export API, handling rate limits, and transforming the JSON payloads to match PostHog’s ingestion schema requires dedicated data engineering pipelines.

Second, to guarantee tracking continuity and validate data parity, organizations must run both SDKs in parallel for at least 30 to 60 days. This parallel-running period generates duplicate ingestion fees and temporary dual-licensing costs.

Third, software developers must swap out Amplitude SDKs for PostHog SDKs across all web, mobile, and server-side codebases. Rebuilding custom event taxonomies, configuring user identity mapping (aliasing), and manually recreating complex behavioral cohorts, funnels, and dashboards in PostHog demands several engineering sprints.

Finally, there is a hidden training cost. Product managers, data analysts, and marketers must adapt to PostHog’s UI and query logic, temporarily lowering team analytical velocity during the transition phase.



Cost and pricing analysis verified as of 2026-06-26. Self-hosting costs are estimates based on standard cloud providers.


よくある質問

How do the scaling costs and MTU limits of Amplitude's Plus tier compare to self-hosting PostHog?

Amplitude's free tier caps at 50,000 monthly tracked users (MTUs), after which you must upgrade to the Plus tier starting at $49 per month (billed annually) or $61 per month, with extra overage fees if you exceed 100,000 MTUs. Because PostHog is an open-source, MIT-licensed platform built on Python, you can self-host it to completely avoid MTU-based subscription caps and rapidly scaling enterprise fees. Furthermore, self-hosting PostHog grants you direct access to product analytics, session recording, and feature flagging without commercial SaaS volume restrictions.

Are features like Session Replay and A/B testing included in Amplitude's base pricing, or is PostHog more cost-effective for these capabilities?

Amplitude treats Amplitude Experiment and Session Replay as hidden costs, billing them separately as paid add-ons outside of the $49/month Plus tier. PostHog, which has a 9/10 overlap score with Amplitude, natively integrates session recording, feature flagging, and A/B testing directly into its core platform. This makes the MIT-licensed PostHog a significantly more cost-effective choice for teams wanting unified qualitative and quantitative analytics without paying compounding add-on fees.

機能と価格データは公式ドキュメントと料金ページを出典としており、最終確認日は 2026年6月26日 です。 誤りを見つけましたか?お知らせいただければ修正します。