アーキテクチャ・意思決定比較表
Intercom と Chatwoot 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。
At 5 agents, Intercom Essential runs $145/month — cheaper than Chatwoot’s self-hosted TCO of $260/month, since infrastructure and DevOps overhead dominate at that scale. The picture flips hard past that point: at 20 agents, Intercom’s Advanced plan plus Fin AI Copilot and 1,000 resolutions climbs to $3,670/month against Chatwoot’s $950/month — a savings of over 74%. At 100 agents, Intercom’s real operational cost (seats, Copilot, 10,000 resolutions) reaches $27,300/month while Chatwoot’s Kubernetes-based TCO stays at $2,700/month, a 90% reduction. Intercom’s usage-based AI fees are what break the budget at scale, not the base seat price.
Intercom’s Official Pricing Structure
Intercom does not offer a free tier. Its pricing is tiered based on feature access and billed per seat, with substantial discounts applied for annual commitments.
| Plan Name | Monthly Price (Per Seat) | Annual Price (Per Seat / Month Equivalent) | Key Included Highlights |
|---|---|---|---|
| Essential | $39 | $29 | Shared inbox, basic workflows, help center access, support for Fin AI Copilot (paid add-on). |
| Advanced | $129 | $99 | Multiple help centers, custom roles, advanced automated workflows, multilingual support. |
| Expert | $169 | $139 | SLA rules, active workload management, native Salesforce integration, priority support. |
Pricing verified as of June 25, 2026.
The Hidden Costs of Intercom
Calculating Intercom’s Total Cost of Ownership (TCO) requires looking far beyond the base seat license. The actual invoice is frequently inflated by secondary fees:
- Fin AI Agent Resolution Fees ($0.99 per resolution): This is Intercom’s autonomous AI bot. Every time the AI successfully resolves a customer issue, you are billed $0.99. For a company handling 5,000 automated resolutions a month, this adds $4,950/month on top of seat licensing.
- Fin AI Copilot Add-on ($35/seat/month): To equip your human agents with real-time draft suggestions and auto-summaries, you must pay an extra $35 per seat, every month.
- API Rate Limiting & Webhooks: High-growth teams often hit standard API limits on lower tiers, forcing them to upgrade to the Expert tier solely for higher rate limit allowances.
- Telephony & SMS Usage Charges: Phone and SMS features are not bundled into base pricing and incur standard carrier fees plus Intercom markup.
Total Cost of Ownership (TCO) Analysis: Chatwoot
Chatwoot is an enterprise-grade, open-source customer engagement suite written in Ruby and Vue. It offers a direct alternative to Intercom’s shared inbox, live chat widget, and omnichannel integrations. Because it is open-source, the software license itself is free, but self-hosting incurs infrastructure and engineering maintenance costs.
Cost Trade-offs at a Glance
1. Hosting & Server Resource Estimation
- Small Team (5 users, <10k monthly visitors): A single VPS instance (e.g., AWS EC2
t3.mediumor a DigitalOcean Droplet with 2 vCPUs, 4GB RAM) hosting Chatwoot, PostgreSQL, and Redis. - Medium Team (20 users, ~50k monthly visitors): Multi-container deployment (AWS ECS or Docker Compose) with dedicated managed services: AWS RDS PostgreSQL (db.t3.medium), managed Redis, and an S3 bucket for media storage.
- Large Team (100+ users, >250k monthly visitors): Highly available Kubernetes (EKS) cluster with auto-scaling application nodes, multi-AZ RDS PostgreSQL cluster, and Elasticache Redis.
2. Maintenance & Engineering Support Estimation
While the software is free, your engineering team must maintain it. This includes applying security patches, orchestrating database backups, scaling resources, and updating Chatwoot versions.
- Small: ~2 hours/month of DevOps oversight ($100/hr internal rate = $200/month).
- Medium: ~6 hours/month ($600/month).
- Large: ~15 hours/month of dedicated DevOps attention for scaling, pipeline integrations, and high-availability testing ($1,500/month).
3. Comparative TCO Table (Monthly Costs)
| Team Scale | Chatwoot Infrastructure (AWS) | Chatwoot Eng. Maintenance | Total Chatwoot TCO | Intercom Base (Annual Rate) |
|---|---|---|---|---|
| Small (5 Seats) | $60 / mo | $200 / mo | $260 / mo | $145 / mo |
| Medium (20 Seats) | $350 / mo | $600 / mo | $950 / mo | $1,980 / mo |
| Large (100 Seats) | $1,200 / mo | $1,500 / mo | $2,700 / mo | $13,900 / mo |
Scenario Comparisons: Intercom vs. Chatwoot
Scenario A: The 5-User Support Team (SaaS Startup)
- Intercom (Essential Plan, Annual Billing): 5 seats × $29 = $145 / month ($1,740 / year).
- Chatwoot (Self-Hosted): Infrastructure ($60) + DevOps overhead ($200) = $260 / month ($3,120 / year).
- Financial Verdict: At small scales, Intercom is more cost-effective. The operational overhead of deploying and maintaining self-hosted infrastructure outweighs Intercom’s base subscription costs.
Scenario B: The 20-User Customer Support Department
- Intercom (Advanced Plan, Annual Billing): 20 seats × $99 = $1,980 / month ($23,760 / year). If the team uses Fin AI Copilot on all seats ($35 × 20) plus 1,000 automated AI resolutions ($0.99 × 1,000), the monthly bill climbs to $3,670 / month ($44,040 / year).
- Chatwoot (Self-Hosted): Dedicated AWS resources ($350) + DevOps support ($600) = $950 / month ($11,400 / year).
- Financial Verdict: Chatwoot saves over 50% (and up to 74% with AI usage). At 20 users, Chatwoot represents a minimum annual savings of $12,360 on base seats alone, rising to over $32,000 when accounting for Intercom’s AI add-on pricing.
Scenario C: The 100-User Enterprise Operation
- Intercom (Expert Plan, Annual Billing): 100 seats × $139 = $13,900 / month ($166,800 / year). Adding Fin AI Copilot ($35 × 100) and an average of 10,000 automated bot resolutions ($0.99 × 10,000) results in an actual operational cost of $27,300 / month ($327,600 / year).
- Chatwoot (Self-Hosted Kubernetes): Enterprise cluster hosting ($1,200) + dedicated DevOps hours ($1,500) = $2,700 / month ($32,400 / year).
- Financial Verdict: Chatwoot saves up to 90% at scale. The enterprise saves nearly $300,000 annually. Engineering leads can also utilize open-source APIs to plug in their own self-hosted models or cost-effective LLM endpoints (such as Claude 4.8 Sonnet or GPT-5.5) via Chatwoot webhooks, avoiding Intercom’s $0.99/resolution markup entirely.
When Does Paying for Intercom Actually Save Money?
Chatwoot’s infrastructure-to-licensing cost advantage at scale is clear, but opting for Intercom’s premium SaaS model is still financially and operationally justified under specific conditions:
- Zero DevOps Bandwidth: If your engineering pipeline is entirely occupied by core product delivery and cannot afford to lose even 5 hours a month to self-hosted helpdesk maintenance, paying Intercom’s premium saves internal engineering cycles.
- Complex CRM and ERP Integrations: If your support flow relies on complex, two-way, out-of-the-box integrations with Salesforce, Marketo, Jira, or custom enterprise systems, building and maintaining these connectors manually in Chatwoot will easily surpass the cost of Intercom’s Expert tier licensing.
- Strict Enterprise Compliance Requirements: If your organization requires instant SOC 2 Type II, HIPAA, or ISO 27001 data compliance at the software level, Intercom provides this natively. Securing, auditing, and certifying a self-hosted Chatwoot instance to meet these standards can cost tens of thousands of dollars in compliance consulting.
Final Purchasing Recommendation
For Financial Planners
If your customer support organization has fewer than 10 agents, stick with Intercom’s Essential plan. It keeps overhead low and allows your lean team to focus on customer acquisition.
However, if your organization has more than 15 agents, or is highly focused on deploying automated customer support bots, mandate an evaluation of Chatwoot. The cost inflection curve for Intercom’s seats and AI resolution fees is extremely steep. Transitioning to Chatwoot caps your customer engagement costs to predictable server fees, saving your bottom line tens of thousands of dollars annually.
For Engineering Leads
If you prioritize data sovereignty, deep customization, and control over where your customer data resides (especially relevant under GDPR and CCPA), Chatwoot is the superior technical path. It runs cleanly on Docker and Kubernetes, and its Ruby/Vue stack is easy to extend.
If your team has the infrastructure capacity, deploying Chatwoot allows you to bypass expensive AI seat add-ons. You can construct a highly efficient, customizable AI agent internally using modern models like Claude 4.8 Haiku or Claude 4.8 Sonnet, routing them into Chatwoot via webhooks for a fraction of Intercom’s $0.99 per resolution fee.
Cost and pricing analysis verified as of 2026-06-25. Self-hosting costs are estimates based on standard cloud providers.
よくある質問
When calculating the total cost of Intercom's AI features, what hidden fees will I face compared to self-hosting Chatwoot under its AGPL-3.0 license?
While Chatwoot can be self-hosted for free under the AGPL-3.0 license, Intercom's pricing scales rapidly with a $35 per seat/month fee for the Fin AI Copilot addon and $0.99 per successful resolution for the Fin AI Agent. This is in addition to Intercom's base seat pricing, which starts at $29 per seat/month annually for the Essential tier and goes up to $139 per seat/month for the Expert tier.
Do I need Intercom’s $99/month Advanced tier for multilingual support and workflows, or does Chatwoot cover these capabilities?
Intercom gates advanced workflows and multilingual support behind its Advanced tier ($99 per seat/month billed annually), while its highest Expert tier ($139 per seat/month) is required for SLA rules and Salesforce integration. With a 9/10 overlap score, Chatwoot's Ruby/Vue-based open-source platform replicates Intercom’s core shared inbox and live chat experience without these restrictive pricing tiers.
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25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。