
Lead Qualification Router
Qualification form → phone-match router → Pipedrive deal → Google Sheets log → "Qualified" / "Not Qualified" tag in ActiveCampaign, before a rep opens the inbox.
Read the case →Home › Blog › Make.com vs n8n for a 5-person team
// blog
I build on both. The contact form on this site runs through a self-hosted n8n instance; the lead funnels in my case studies run on Make.com. Neither is "better" — they fail in different places, and a small team should pick by the three questions below, not by feature lists.
Make.com when the people who will maintain the automation are operations staff, your volume fits a standard plan, and every app you use has a native module. n8n when data must stay on your server, you run enough volume that per-operation pricing hurts, or the flow needs real code in the middle. If you are unsure, Make.com is the lower-risk first choice — migrating later is a known procedure, not a rewrite.
| Make.com | n8n | |
|---|---|---|
| Pricing model | Per operation, monthly plan; every module run counts | Cloud plan per execution, or self-hosted — then the server is the only cost |
| Hosting | Managed; nothing to patch | Cloud (managed) or your VPS in Docker — you own updates, backups, HTTPS |
| Editor | Visual, polished; routers, iterators, aggregators as first-class blocks | Visual; expressions are JavaScript, Code node for anything else |
| Custom logic | Functions inside mapping; no arbitrary code | Full JavaScript/Python nodes, custom HTTP with pagination |
| Error handling | Error handlers per module: retry, resume, ignore, rollback | Per-node retry + a global error workflow (I ship this one) |
| Integrations | ~2,000 apps, consistent quality | ~1,000 nodes plus community nodes of varying quality; HTTP node covers the rest |
| AI steps | OpenAI / Claude modules | Native AI agent and LangChain-style nodes, tools, memory |
| Who fixes it at 9 am Monday | Anyone who can read the canvas | Someone comfortable with expressions and a server |
n8n, self-hosted: the form on this site posts to a webhook; a filter drops empty bot submissions; three branches run in parallel — a Telegram alert, a row in Google Sheets, an email copy. Cost: one small DigitalOcean droplet that also hosts other things. The export is free to import.
Make.com: a retailer's lead funnel — four scenarios across ClickFunnels, Cal.com, Pipedrive and ActiveCampaign, maintained by the client's own team after handoff. The editor they can read and the per-module error handlers mattered more than the per-operation price. See Lead Qualification Router and Booking Cancellation Recovery.
From Make.com to n8n: when the monthly bill crosses what a server plus a day of my time costs, or when a data rule appears. From n8n to Make.com: when the one person who understood the server leaves. Both moves are inventory → rebuild with error handling → parallel run → cutover; the migration page describes the procedure (it is the same one for Zapier).
Start on Make.com with the two flows that eat the most hours, keep every scenario documented and in the client's organisation, and revisit the platform question at the first invoice that makes you wince. If any of the three questions above already has a hard answer, skip the experiment and go straight to n8n.

Qualification form → phone-match router → Pipedrive deal → Google Sheets log → "Qualified" / "Not Qualified" tag in ActiveCampaign, before a rep opens the inbox.
Read the case →
Cancelled Cal.com call → Pipedrive lookup → cancellation logged → "Call Discovery Cancel" tag → deal moved back a stage so the lead gets a real follow-up.
Read the case →Website form → HubSpot CRM → Telegram alert to sales → personalised email auto-reply. Zero manual steps.
Read the case →Which tools are involved, what is done by hand today, how often. I reply within 24 hours with a fit/no-fit answer and, if it fits, a fixed quote.