Automation
n8n, Make or Zapier: which should a UK small business use?
By Felix Grant5 min read
Zapier is the easiest for non-technical teams to understand and edit, and gets expensive at volume. Make handles more complex workflows for less money, with a steeper learning curve. n8n is the most flexible and can run on your own server, but it needs someone technical to look after it. For most small businesses, the right choice depends less on features than on who will maintain the automation after it is built.
Compare
The short comparison
Zapier suits non-technical teams and simple, linear workflows. You build it as a step by step list, it runs only on Zapier's cloud, and most office staff can maintain it after a short handover.
Make suits multi-step workflows with branches and data shaping. You build on a visual canvas, it runs on Make's cloud, and a confident spreadsheet user can maintain it after some training.
n8n suits complex workflows, custom code, AI agents and self-hosting. You build on a visual canvas with code where needed, it runs on n8n's cloud or your own server, and it needs someone technical to look after it. Pricing plans change often, so check each vendor's current pricing page before you decide.
Pricing
How each prices, and why it matters
The pricing models matter more than the headline monthly fee, because they decide what happens as you grow. Zapier charges by task, so a workflow with five actions that runs 1,000 times a month uses 5,000 tasks. It is simple to understand, and the cost rises quickly with volume or long workflows.
Make charges by operation, roughly each module that runs. It is usually cheaper than Zapier for the same work, especially for workflows that process lists or many records. n8n charges by workflow execution on its cloud plans, so one run of a long workflow counts once. Self-hosted, there is no per-run fee, but you pay for the server and for someone to look after it.
The practical test is to estimate how many times each workflow will run each month and how many steps it has, then price it on each platform. A workflow that is cheapest on Zapier at 100 runs can be the most expensive at 10,000.
Hosting
Hosting and data location
For most small businesses, cloud hosting on any of the three is fine. It matters more when you handle sensitive personal data and need to control where it is processed, when a client or contract requires data to stay in the UK or EU, or when you want to keep credentials for internal systems off a third-party platform.
Zapier and Make run on their own cloud infrastructure, so check their current data residency options and data processing terms. n8n can run on a server you choose, including in the UK, which gives you the most control and the most responsibility.
Maintenance
Maintenance, and who can edit it later
Automations break. An API changes, a password expires, a supplier renames a field, a form adds a question. What matters is how quickly you find out and how easily you can fix it. Zapier shows failed runs clearly and emails you, and fixes are usually simple because workflows are simple. Make gives detailed run histories and good tools for retrying and handling errors on specific steps. n8n offers the most control over error handling, but someone has to set it up, and self-hosted you also own updates, backups and uptime.
This is the question most comparisons skip, and the one we ask first. If your own team will make changes and nobody is technical, Zapier is usually the safer choice even if it costs more each month. If one person is comfortable with spreadsheets and logic, Make is often the best balance. If a developer or a partner will maintain it, n8n gives the most room to grow.
Whichever you pick, every automation that touches money, customers or compliance should alert a named person when it fails, and you should get a written handover covering what each workflow does, what triggers it, what it connects to and how to turn it off.
Choose the platform your team can keep running.
Costs
What UK agencies charge to build on each
A study of 277 UK AI automation agencies by TrustedAIAgencies, reported by IT Brief, found that 59% publish no prices at all. Among those that do, the median starting price was £875 for an n8n project, £950 for Make and £975 for Zapier. A general workflow automation build started at £1,500, an AI agent or custom build at £4,000, an automation audit at £750, and a monthly plan at £250.
These are starting prices for a single simple workflow. Anything that touches several systems, needs error handling or involves customer data costs more. Our Build package starts at £2,500 for one scoped workflow connecting up to three systems, including testing, failure alerts and a written handover.
Our choice
Our default choice, and why
We choose Zapier when the client's team will own and edit it, the workflows are simple and volumes are low. We choose Make when workflows need branching, loops or data shaping and someone on the client side can learn the canvas. We choose n8n when we will maintain it under a Partner arrangement, when data needs to stay on a server the client controls, or when the workflow uses AI steps, custom code or many systems.
We use n8n with AI steps for most of our own builds, such as invoice chasing, reporting and content publishing with a veto window. Moving platforms later is possible, but workflows have to be rebuilt rather than imported, which is another reason to document them.
If you are not sure which fits, a Discover workshop at £950 maps the workflows worth automating, estimates run volumes, and recommends the platform your team can actually maintain.
Key takeaways
- Zapier for simple workflows your own team will edit
- Make for complex workflows at lower running cost
- n8n for flexibility, AI steps and control over where data lives
- Price each workflow on runs and steps, not the headline plan
- Pick the platform your team can maintain after the builder leaves

Written by
Felix Grant
Founder of Kitebox, a business technology and AI consultancy that builds what it recommends. Based in Mallorca, working with UK businesses.
About Felix →