10 Apps Built With n8n That Have Revenue Evidence

Apps built with n8n make money in this cohort, but the strongest evidence favors services and distribution systems, not a magic passive-income workflow. ProvenStartups found 10 projects, seven solo-run; only three publish a clean monthly figure, with a $38K/mo [C] median across the full matching set. The practical call is to sell a narrow business outcome first and keep n8n behind the curtain.

Contents

• ·The 10 revenue-linked n8n projects
• ·What the full n8n cohort actually shows
• ·Where the data contradicts n8n hype
• ·What we would build, and refuse to build
• ·How to read the revenue evidence
• ·FAQ

The 10 Revenue-Linked n8n Projects

These are all 10 projects in the matching cohort, not a hand-picked winners list. The useful comparison is the revenue unit: recurring software revenue, project fees, content revenue, or sales. Several rows prove that someone earned money near an n8n workflow; they do not prove the workflow itself was the product.

A grade describes the source of the claim, not whether the business is attractive. [F] means founder-reported; [C] means creator-relayed. “Revenue attached” can still mean a launch day, a client project, or a reference business rather than durable MRR.

What the Full n8n Cohort Actually Shows

The full matching set is small, solo-heavy, and economically uneven. Seven of the 10 projects are solo-run, while only three disclose a clean monthly result. Across that full cohort, not merely selected examples, the median is $38K/mo [C] and the range runs from $37/mo [F] to about $400K/mo [F].

The mix is three SaaS projects, three AI services, two cautionary tales, one AI e-commerce case, and one AI content case. That distribution matters more than the headline median: project fees, agency revenue, AdSense, credit consumption, and subscriptions are not interchangeable.

n8n also appears in only 10 of the 211 distinct projects that mention at least one tracked coding tool. For context, ChatGPT appears in 100, Claude Code in 50, and Cursor in 46. n8n is a useful implementation layer here, not the dominant source of startup outcomes.

Where the Data Contradicts n8n Hype

The popular claim is that n8n turns a clever workflow into passive SaaS. This cohort says the opposite: the clearer path is selling implementation, distribution, or a measurable client outcome, often with human delivery still involved. Two cases are explicitly cautionary, and one showcased clone has no revenue attached to the clone itself.

StreamWeaver climbed from websites into automation contracts and reported $124,000 net [F]. True Horizon AI reported $23K [F] across project deals, not retainers. Those are sales-and-delivery businesses powered by automation, not unattended workflow vending machines.

Even the loudest product result needs separation. Minea peaked at $750K MRR [F], but that does not establish n8n as its moat. The contradiction is blunt: workflow assembly is getting easier, while distribution, customer access, and domain-specific execution remain scarce.

What We Would Build, and Refuse to Build

ProvenStartups would build a narrow, paid workflow for a buyer with an expensive recurring problem. Start as a service, observe the repeated steps, then productize only the stable pieces. We would refuse generic “AI automation agency” positioning and faceless-content clones because neither identifies a buyer, a defensible input, or reliable demand.

1. 1.Sell the output. Quote a lead-recovery system, claims intake flow, catalog operation, or reporting process, not “an n8n automation.”
2. 2.Keep a human checkpoint. Use n8n’s official AI agent documentation to design tool access and review points around consequential actions.
3. 3.Price from delivered value. Check n8n’s published pricing tiers, then model execution cost separately from implementation, monitoring, and support.

The e-commerce agents’ roughly $8,000 [F] user-revenue claim is interesting, but much of the demo used hypothetical competitors. That is a validation prompt, not permission to build four agents before speaking to a merchant.

How to Read the Revenue Evidence

Trust the row-level grades, not the blank cohort aggregate. The cohort summary shows zero [V], zero [F], zero [C], and zero [U], yet its case records carry source grades. Those fields conflict, so ProvenStartups preserves the disclosed row-level classification and states exactly what each figure covers instead of laundering the mismatch into certainty.

Across the full index of 406 startup ideas, 57 are third-party verified [V], 184 founder-reported [F], 121 creator-relayed [C], and 44 unverified [U]. The grading method lets readers distinguish a verified figure from a founder statement before copying a business model.

For this page, inspect three things: who made the claim, whether the revenue belongs to the n8n-built version, and whether the unit is MRR, total sales, a project fee, or one-day revenue. Quick Shorts made $500+ [F] on launch day, but ProvenStartups files it as a cautionary tale, not a repeatable win.

FAQ

The short answers are operational: n8n is infrastructure, the evidence is mostly self-reported, and the better entry point is usually a paid service with a narrow outcome. A solo founder can run this model, but the cohort does not support treating a copied workflow, a demo result, or someone else’s channel revenue as validation.

Are these businesses actually built with n8n?

They all match the site’s n8n cohort, but n8n plays different roles. It may orchestrate agents, connect a content pipeline, support client delivery, or sit inside a broader SaaS system. The table therefore says “revenue-linked,” not “revenue caused by n8n.” Read each case page before treating the tool as the product.

Is n8n enough to create a defensible SaaS?

No. n8n can shorten implementation, but accessible workflow tooling also shortens a competitor’s implementation. Defensibility must come from proprietary data, embedded operations, distribution, trust, or switching costs. Leftclick’s case matters because it explicitly argues that automation skill itself is expiring, despite about $400K/month [F] across the founder’s businesses.

Which n8n business model would ProvenStartups choose?

A productized service for one expensive workflow is the strongest starting point. It can earn project revenue before the software boundary is clear, as True Horizon’s $23K [F] deal total illustrates. After repeated delivery, convert only the common components into software; keep custom exceptions priced as service work.

Can one person run an n8n business?

Yes, seven projects in this 10-case cohort are solo-run. That supports operational feasibility, not automatic profitability. A solo founder should constrain integrations, buyer type, and support promises, then reject work that creates bespoke maintenance. Insurance Sales Genie’s $37/mo [F] price shows that a paid niche tool can start small even when subscriber count is not disclosed.