If you have looked for ways to automate processes in your company without coding from scratch, you have probably come across the same three names: n8n, Zapier and Make. All three connect applications, all three have visual drag-and-drop interfaces, and all three promise to save you hours of manual work. However, they work in very different ways, and choosing the wrong one can cost you time, money, or both. This comparison, based on real 2025 data, gives you the criteria to decide clearly.
What does each platform do and what is its philosophy?
Zapier: the no-code Swiss army knife
Zapier is the pioneering platform in no-code automation, founded in 2011. Its key strength is the breadth of integrations: it surpasses 7,000 available connectors as of mid-2025, making it the option with the widest application coverage. The learning curve is minimal: any employee without a technical background can build a basic Zap in under twenty minutes. The trade-off is cost: the pricing model based on executed tasks scales quickly as volumes grow, and more complex automations with advanced conditional logic become cumbersome to maintain in its editor.
Make (formerly Integromat): the visual powerhouse
Make, rebranded in 2022 (the platform was acquired by Celonis in 2020), bets on a workflow editor in diagram form where every step can be seen with precision. It is significantly more powerful than Zapier when handling complex data structures, iterations, array transformations and multi-branch logic. The visual interface is more intuitive for those who need to read and audit elaborate flows. Pricing is measured in operations (each action within a scenario), which can make flows with many nodes more expensive but also allows greater control over spending.
n8n: the sovereign automator
n8n (pronounced «n eight n», founded in 2019) is the only one of the three that offers an open-source model with a self-hosting option. You can download it and run it on your own server without paying a licence fee based on the number of flows or executions. It has more than 400 native nodes and allows you to run JavaScript or Python code directly inside the flow, making it the preferred option when the automation needs custom business logic. The learning curve is steeper than Zapier's but reasonably accessible for mid-level technical profiles. Since 2024, n8n has accelerated the development of its native AI features, including agents that call language models and process documents within the same flow.
Comparison table: n8n vs Zapier vs Make
| Criterion | n8n | Zapier | Make |
|---|---|---|---|
| Licence model | Open source (free self-hosted) + paid SaaS | SaaS (no self-hosted version) | SaaS (no self-hosted version) |
| Free plan | Yes (self-hosted with unlimited executions) | Yes (unlimited Zaps, 100 tasks/month) | Yes (1,000 operations/month) |
| Native integrations | 400+ nodes | 7,000+ applications | 1,500+ applications |
| Custom code | Yes (native JS and Python) | Yes (Code by Zapier, limited) | Yes (HTTP module + functions) |
| Native AI | Yes (agents, LLM, RAG chains) | Yes (Zapier AI, ChatGPT Actions) | Yes (OpenAI, Claude integrated) |
| Learning curve | Medium-high | Low | Medium |
| Estimated cost at 50,000 ops/month | From ~€20/month (cloud) or own server | From ~€69/month (Professional plan) | From ~€16/month (Core plan) |
| Data control (GDPR) | Full with self-hosted; EU region in EU cloud | Data in the US (Privacy Shield) | EU region available |
| Ideal for | Complex flows, AI, data sovereignty | Simple automations, speed | Elaborate visual flows, controlled cost |
Indicative prices from plans published as of April 2025. Always verify current plans on each official website.
Real use cases for Spanish SMEs
Case 1: accounting firm with client onboarding
A tax advisory firm with 20 employees receives new clients through a web form. With Zapier, the form automatically creates the record in the CRM, sends the welcome email and opens a task in the project manager. The flow is set up in an afternoon without technical help. For this need, Zapier is the right answer: fast, frictionless and with a reasonable monthly cost for the volume of a small practice.
Case 2: distribution company with order synchronisation
A mid-sized distributor needs to synchronise orders between its online store (WooCommerce), its ERP (Odoo) and its logistics platform. The flow involves transforming data structures, iterating order lines and handling errors when stock does not match. Here Make shines: its scenario editor lets you see the complete flow map, debug each node and handle product arrays with native iterators. The cost per operation is competitive for this volume.
Case 3: professional services company with integrated AI
A consultancy wants to automate the classification and preliminary response of support emails: the flow extracts the email text, sends it to a language model, classifies the query and drafts a response that goes through human review. With n8n on an own server, client data never leaves the company's infrastructure, GDPR compliance is achieved without additional agreements with SaaS platforms and the cost of executions is practically zero beyond the server. For flows involving AI and data sovereignty requirements, n8n is the most complete platform in 2025.
Case 4: online store that notifies suppliers
A small e-commerce business needs to send an email to its supplier every time a product exceeds a sales threshold. The fastest and most economical solution is a simple two-step Zap. There is no reason to deploy Make or n8n when the need is linear and the volume is low.
Which platform does a Spanish SME choose in 2025?
The answer depends on three factors: technical complexity of the flow, monthly operation volume and data privacy requirements. At Summum we work with all three, and the guideline we follow is:
- Zapier when the goal is to set something up in hours and the flow is linear. Companies without internal technical profiles that need quick results for low-complexity automations.
- Make when the flow requires data transformations, complex branches and you want to visualise the complete map. A good balance between power and cost for SMEs of 20 to 100 employees with some technical profile.
- n8n when automation is combined with generative AI, when data cannot leave the organisation, or when execution volume would make SaaS model costs prohibitive. It is the platform we deploy most frequently in advanced automation projects precisely for its flexibility and the total control it offers over data.
The GDPR factor: where does your data live?
For many Spanish SMEs, this point is decisive. Zapier stores execution data on servers in the United States under the EU-U.S. Data Privacy Framework (in force since July 2023), which is legally valid but introduces a dependency on regulatory decisions that may change. Make allows you to select the EU region for data storage. n8n self-hosted keeps everything within the company's own infrastructure, which greatly simplifies GDPR documentation and risk analysis when flows process personal data of clients or employees.
If your automation flows will handle payroll, medical records, data about minors or commercially sensitive information, n8n's self-hosted option significantly reduces the compliance surface area. In our n8n automation service we include environment configuration, credential management and processing records documentation to comply with Regulation (EU) 2016/679.
Real cost at scale: the task and operation trap
Zapier's pricing model charges per task (each individual action in a Zap). A five-step flow that executes 10,000 times a month consumes 50,000 tasks. With the Professional plan (approximately €69/month in 2025 for a mid-range task volume, with additional cost for overages), volumes grow quickly. Make charges per operation with similar logic but with a slightly lower price per operation and more granular plans. n8n in the cloud charges by the number of active flows or workflows on its starter plan (approximately €20/month for a small team), with no execution limits; on self-hosted, the cost is that of the server, which for an SME can be between €10 and €30 per month on a standard VPS.
For companies with more than 20,000 monthly executions of multi-step flows, n8n self-hosted is typically two to five times more economical than equivalent Zapier or Make plans.
Frequently asked questions
Can I use n8n without knowing how to code?
Yes, with some nuance. n8n has a visual node interface and many flows are built without writing a single line of code. However, when you need to transform data with custom logic or handle complex JSON structures, basic JavaScript knowledge speeds things up considerably. For an SME without an internal technical profile, the most common approach is to start with a consultant who sets up the flows and leaves the platform documented and ready so the team can make minor adjustments.
Is Zapier still the best option for simple automations?
For two or three-step automations with popular applications (Gmail, Google Sheets, HubSpot, Slack, Notion), Zapier remains the fastest to implement. Speed of deployment is its greatest advantage. The problem arises when flows grow in complexity or the volume of executions drives up monthly costs: at that point, migrating to Make or n8n is usually the most cost-effective decision.
What happens if the SaaS platform goes down?
Zapier and Make are cloud services and their incidents affect all clients simultaneously. Both have documented incident histories on their status pages. n8n self-hosted depends on the availability of your own infrastructure, which shifts the responsibility but also the control. For critical flows (automatic invoicing, stock alerts, client integrations), it is worth evaluating the retry and alert mechanisms offered by each platform, as all three have some type of error management and notifications.
Is Make the same as Integromat?
Yes. Make is the current commercial name of the platform that was called Integromat until 2022. The company was acquired by Celonis in 2020 and rebranded as Make in 2022. The core functionality is the same, although since the name change they have added AI features, editor improvements and new connectors. Integromat accounts were automatically migrated to Make.