Zapier, Make or n8n for Kommo: How to Choose
If you are automating Kommo and asking this question, you have already outgrown the built-in widgets. The short answer: Zapier makes sense when your team is non-technical and you need to go live in a day; Make - when the logic is more complex than a single branch and the budget is tight; n8n - when task volume is high, data must not leave your infrastructure, or you are already paying Zapier more than $200/month and watching the spend. None of the three covers scenarios with custom business logic at the Kommo data layer - that requires custom development.
This article is written for teams already working with Kommo who are choosing a tool for a real integration, not for people still learning what a CRM is. Prices are current as of June 2026.
Quick Verdict
| Criterion | Zapier | Make | n8n |
|---|---|---|---|
| Price at scale | Expensive: from $20/month for 750 tasks, $49/month for 2,000 | Moderate: from $9/month for 10,000 operations | Cheap: self-hosted from $3-7/month on a VPS, cloud from 20 EUR/month for 2,500 executions |
| Logic complexity | Linear chains, basic branches | Visual scenarios with routers, iterators, aggregators | Full execution graph, JavaScript nodes, conditional branches of any depth |
| Self-hosted | No | No | Yes, Community Edition - free, no limits |
| Idempotency | Limited (no built-in deduplication) | Basic (filters), no built-in deduplication | Configurable via custom logic; requires manual implementation |
| Kommo support | Official connector, limited triggers | Official module, more actions | No official connector - HTTP Request or community node only |
Important note on Kommo: none of the three platforms has a full two-way connector for Kommo covering the entire API. The official Zapier and Make modules handle basic CRUD operations (create a lead, update a field, add a contact), but they do not expose Digital Pipeline events, do not fully support activity note operations, and do not handle Kommo Webhooks in complex configurations. For more on the limits of native integrations, see Custom Integrations for Kommo CRM.
Zapier - Who It Is For and When
When Zapier Makes Sense
Speed to launch is the main advantage. If you need to connect Kommo to Gmail, Slack, and Google Sheets in a day without a developer, Zapier will do it. The official Kommo connector is in the catalog, the drag-and-drop interface is approachable for non-technical teams.
Ecosystem - 7,000+ integrations. If your stack includes a niche B2B tool (a specialized logistics system, a regional bank payment gateway, an industry-specific ERP), the chances that only Zapier has it are high. Make and n8n fall short on catalog breadth.
Low task volume - at 500-1,500 tasks/month Zapier is quite competitive on price. The Professional plan starts at $20/month (billed annually) for 750 tasks. If you are running 10-15 Zaps at low frequency, this is a reasonable choice.
When Zapier Becomes a Problem
Per-task billing hits hard with complex Zaps. Every step in a multi-step Zap counts as a separate task. A five-step Zap with 1,000 triggers/month = 5,000 tasks. On the $49/month plan (2,000 tasks) that means exceeding the limit and triggering automatic overage billing at roughly 1.25x the base per-task rate.
No self-hosted option. All data flows through Zapier servers in the US. For companies with GDPR requirements or internal data residency policies, this is a hard blocker.
Logic is limited. Paths exist, but nested conditions and dynamic loops are unwieldy to build. Error handling is basic: either the Zap stops or it continues. There is no native retry with exponential backoff.
Zapier bottom line: the right choice for a non-technical team with low volume and niche connectors. The wrong choice when volume grows or data residency matters.
Make - Who It Is For and When
When Make Makes Sense
Visual logic without code is Make’s strongest point. A scenario with a router, multiple branches, an iterator over an array, and an aggregator for the results can be built visually and reads like a flowchart. The same thing in Zapier requires multiple separate Zaps or an expensive plan with Paths.
Price at scale. Core plan: $9/month for 10,000 operations. Pro plan: $16/month for 10,000 operations (with priority execution and full logs). That is 3-5x cheaper than Zapier at comparable volumes. Note that Make recently moved to a credit model - most operations cost 1 credit, but AI steps can cost several credits: factor this in for scenarios with GPT integrations.
The official Kommo module in Make is broader than the one in Zapier: more available actions, webhook trigger support, ability to work with custom fields. For standard integration scenarios this is sufficient.
When Make Becomes a Problem
Operations billing on errors. If a scenario fails halfway through, the operations up to the failure point are already consumed. With unstable external APIs (third-party services with downtime) this creates an “operations leak.” You need error handling on every HTTP call.
No self-hosted option. Like Zapier, Make runs only as a cloud service. Data is stored in Make’s infrastructure (servers in the EU, but under Make’s control, not yours).
Complex data transformations are a pain point. JSON parsing, building dynamic requests, working with nested structures is possible but requires knowing Make’s syntax. A newcomer without preparation will get stuck.
Make bottom line: the optimal choice for teams that need visual logic, a moderate technical level, and a lower budget than Zapier. Falls short when self-hosted is required or when high-frequency scenarios involve unstable upstream services.
n8n - Who It Is For and When
When n8n Makes Sense
Self-hosted with no execution limits. n8n Community Edition is a fully free open-source release. A VPS at $5-7/month (DigitalOcean, Hetzner) gives you a full n8n instance with unlimited workflows and executions. This is a fundamentally different cost structure at volumes of 5,000+ executions/month.
Full data control. Scenarios run on your server; data never leaves your infrastructure. For companies with GDPR requirements, internal compliance policies, or simply an unwillingness to send CRM data to a third party, this is the decisive argument.
Logic power. JavaScript nodes allow arbitrary data transformations directly inside the workflow. The execution graph supports any depth of nesting, merge nodes for combining streams, and split nodes for parallel execution.
AI capabilities (n8n 2.0, December 2025). 70+ AI nodes with LangChain integration. If you are building AI agents on top of Kommo (lead classification, call summarization, auto-replies), n8n provides the most flexible platform at no extra charge for AI steps.
When n8n Becomes a Problem
No official Kommo connector. n8n has no native Kommo node. All integration is built through HTTP Request nodes and the Kommo REST API manually. That means: you need to know the Kommo API, handle OAuth2 authentication yourself, and build payloads by hand. Routine for a developer; a barrier for a non-technical team.
DevOps required for self-hosted. Installation, updates, backups, uptime monitoring - all on you. The n8n Cloud version (Starter: 20 EUR/month for 2,500 executions) removes that burden, but it is no longer cheaper than Make at comparable volumes.
Learning curve. n8n’s interface is more powerful but also more complex. A team without a technical background will spend considerably more time on their first working workflow.
n8n bottom line: the right choice for technical teams with high volume or data residency requirements. The wrong choice for a non-technical team without a DevOps resource.
Real Case: Migrating from Zapier to n8n and How Much Was Saved
One Exceltic.dev client - a SaaS company with an 8-person sales team working through Kommo. Their automation stack included:
- Syncing new leads from website forms into Kommo (webhooks -> Kommo API)
- Slack notifications on pipeline stage changes
- Creating tasks in ClickUp when a lead moved to the “Proposal” stage
- A weekly conversion report in Google Sheets
Total volume: around 8,000-9,000 Zapier tasks/month (multi-step Zaps). They were paying $99/month for the Team plan (2,000 tasks + overage). The actual bill with overage came to $140-160/month.
After migrating to self-hosted n8n (Hetzner CX21 VPS, 5.50 EUR/month):
- Infrastructure cost: 5.50 EUR/month
- Migration time: 3 developer working days (VPS setup, workflow logic transfer, testing)
- Savings: approximately $130-155/month, i.e. $1,560-1,860/year
One important detail: some Zaps used connectors that do not exist in n8n (one niche B2B service). For that connector they kept Zapier on the minimum plan ($20/month). Net savings came to roughly $110-130/month.
The migration paid for itself in 3-4 weeks. Without a developer on the team the payback period stretches out, and in that case Make may be a more practical intermediate option.
Break-even calculation: if your current Zapier bill exceeds $60/month and volume keeps growing, compare the cost of Make Pro ($16/month) or n8n self-hosted ($5-7/month on a VPS) against the time spent on migration. At a developer rate of $50/hour and 16 hours of migration work, break-even arrives around month 6 when saving $80/month.
When None of the Three Works: Custom Development
Zapier, Make, and n8n are tools for orchestrating data between APIs. They work well when both ends of the integration have a stable REST API with a predictable schema. Kommo largely meets that requirement - but with caveats.
Scenarios where automation platforms will not get the job done:
Complex business logic at the Kommo data layer. For example: when a deal is created, check whether a contact with that email already exists, merge duplicates, transfer the activity history from the old deal, and reassign the owner according to territory logic. This is not “connect two APIs” - it is transactional logic with rollback on error. Zapier/Make/n8n will not handle this reliably.
Two-way sync with an external system. If changes in Kommo need to be reflected in an ERP and vice versa, without conflicts and with collision resolution, you need a stateful layer that stores sync state. Automation platforms are stateless by nature.
Non-standard Kommo triggers and events. Digital Pipeline events, changes in activity notes, working with custom widgets - none of this is covered by standard connectors. You need a webhook handler with custom event-parsing logic.
High reliability with an SLA. If a missed execution is a business loss (a missed lead, an unsent document, an uninvoiced order), you need a system with monitoring, retry logic, alerting, and delivery guarantees. That is built, not configured in a UI.
For more on when native integrations are not enough, see Custom Integrations for Kommo CRM. For an overview of the platform as a whole, see Kommo CRM Overview.
Frequently Asked Questions
Is there an official Kommo connector in n8n?
No. The official n8n catalog does not include a native Kommo node. Integration is built through the “HTTP Request” node with manual configuration of the Kommo REST API and OAuth2 authentication. This works, but requires technical knowledge. Unofficial solutions exist in the community node catalog, but their quality and ongoing support are not guaranteed.
How are operations counted in Make vs tasks in Zapier?
In Zapier, every successful step in a Zap = 1 task. A five-step Zap = 5 tasks per execution. In Make, every processed data item on every module = 1 operation (now 1 credit in the new model). For the same scenario, Make typically consumes fewer billing units than Zapier, and the base limit is 10x higher at a comparable price point.
Can I use n8n Cloud instead of self-hosted?
Yes. Cloud Starter: 20 EUR/month for 2,500 executions; Cloud Pro: 50 EUR/month for 10,000 executions. This removes the need for a DevOps resource but eliminates n8n’s main pricing advantage. At up to 10,000 executions, n8n Cloud Pro (50 EUR/month) is comparable to Make Pro ($16/month for 10,000 operations) - Make ends up cheaper for cloud deployments. n8n Cloud makes sense if you need AI nodes or specific n8n capabilities but have no DevOps resource.
What is cheaper at 20,000 executions/month?
Zapier: you need a plan around $100/month or more (Team). Make: you will need to buy additional operations or upgrade to a higher tier - roughly $20-30/month. n8n self-hosted: $5-7/month for the VPS regardless of volume. n8n Cloud Business: 667 EUR/month for 40,000 executions - not justified at that volume, self-hosted is clearly better.
Does Kommo support incoming webhooks for Zapier and Make?
Yes, Kommo can send webhooks on events (deal created, stage changed, new contact). Both Zapier and Make can receive these webhooks as a trigger. However, the Kommo event format is specific, and not all fields are available out of the box - field mapping needs adjustment. For more complex events (Digital Pipeline actions, widget changes), webhook triggers in the standard connectors do not work.
Final Decision Matrix
Choose Zapier if: your team is non-technical, you need niche connectors not available on other platforms, and volume is under 1,500-2,000 tasks/month.
Choose Make if: you need visually complex logic, the team has a basic technical level, cloud hosting is acceptable, and budget is a constraint.
Choose n8n self-hosted if: you have a developer or DevOps resource, volume is high (5,000+ executions/month), and data must not leave your infrastructure.
None of the three fits if you need transactional business logic, two-way sync with conflict resolution, or guaranteed delivery with an SLA. In those cases, custom development is the answer.
For a comparison of all Kommo automation options - including native widgets and API integrations - see the Integrations and Development section.
Prices current at time of writing (June 2026): zapier.com/pricing, make.com/en/pricing.