10 n8n Workflow Examples You Can Actually Build and Sell

Most of the n8n workflows people post to look clever are four nodes and a webhook. Trigger, HTTP request, Slack message, applause. They demo beautifully and solve nothing a business would pay to keep running.

The workflows worth stealing are the boring ones. The ones that quietly move data between tools a company already pays for, every hour, while nobody watches. Those turn into a monthly invoice.

I've watched a lot of people learn n8n, build one flashy AI demo, and then freeze because they can't connect it to a real problem. So here are ten that map to money. You'll get the node chain for each, and my honest read on who actually pays.

If you're brand new to the tool, start with what n8n is and the step-by-step n8n tutorial first, then come back for the ideas.

One line to carry through the whole thing: a workflow is only worth building if someone would notice when it stops.

Marketing workflows

Marketers are the easiest sale, because they already know that copy-pasting all afternoon is time they didn't spend on strategy.

1. Lead capture, enrich, and route (beginner). A form lead comes in, gets cleaned up, gets scored, and lands where sales actually looks.

Node chain: Webhook (form submission) -> HTTP Request (enrichment API like Clearbit or Apollo) -> IF/Switch (route by company size or budget) -> HubSpot/Airtable node (create contact) -> Slack node (alert the owner).

Who pays: any B2B service business drowning in Typeform submissions nobody follows up on fast enough. $300-600/mo, and you can prove it, because speed-to-lead is one of the few things a client can measure without you explaining the math to them.

2. Abandoned-form and abandoned-cart follow-up (intermediate). Someone starts a checkout, bails. The flow waits, checks whether they finished, nudges them if they didn't.

Node chain: Webhook (form/cart started) -> Wait node (30-60 min) -> HTTP Request (check if the order or submission completed) -> IF (not completed) -> Email/SMS node (send the nudge).

The first version of this I built emailed a guy who had already bought, because the completion check was reading a stale order status and my Wait was too short for their payment processor to settle. He was polite about it. His agency owner was not. Set the wait longer than you think, and re-check the order right before you send, not before you wait.

Who pays: e-commerce stores and course sellers. Recover 5% of abandoned carts and your retainer pays for itself ten times over. Easy build, but the tracking has to be honest.

3. AI content pipeline: news to draft social posts (intermediate). Pull fresh articles in the client's niche, summarize, draft posts in their voice, queue everything for a human.

Node chain: RSS Feed / Schedule Trigger -> HTTP Request (fetch full article) -> AI/LLM node (summarize + draft 3 posts) -> Airtable or Notion node (save as "pending approval") -> Slack node (ping the marketer to review).

Who pays: personal brands, founders, small marketing teams. Comfortable $500-1000/mo. The approval step is the whole trick, and I'd argue it's the reason the flow survives past month two. One hallucinated stat published to 40,000 followers and the relationship is over.

Sales and CRM workflows

Sales teams buy anything that takes admin off the day, because admin is what keeps them off the phone.

4. New customer onboarding across apps (intermediate). Deal closes, customer gets set up everywhere, nobody has to remember the checklist.

Node chain: Stripe/CRM Trigger (payment or "won" stage) -> Create records in parallel: Google Workspace or app account, project in Asana/ClickUp, contact in the email tool -> Email node (branded welcome) -> Slack node (notify the account manager).

Who pays: SaaS companies and agencies onboarding by hand. Every new customer is 30-45 minutes of clicking. Multiply by their monthly signups and you have the number that goes on the invoice.

5. Support ticket triage with AI tagging (advanced). Every ticket gets read, categorized, prioritized, and routed before a human touches it.

Node chain: Email/Help-desk Trigger (Zendesk, Freshdesk, Gmail) -> AI/LLM node (classify: category + urgency + sentiment) -> Switch (route by category) -> update ticket + assign -> Slack node (escalate anything flagged angry or urgent).

It's advanced because the classifier has to be tuned to their actual product, not to a generic taxonomy you copied from a template. Budget a week of watching it get things wrong. But once it runs, nobody wants to go back, and that stickiness is what you're really selling.

Operations and finance workflows

The least glamorous work on this list. Which is exactly why the money is good and almost nobody is competing for it.

6. Invoice and receipt data extraction (advanced). PDF invoice lands in an inbox, the numbers come out, nobody types them.

Node chain: Email Trigger or Google Drive Trigger (new PDF) -> Extract From File node (get the text) -> AI/LLM node (pull vendor, amount, date, line items into structured JSON) -> Google Sheets or QuickBooks node (append the row) -> Slack node (flag anything the AI wasn't confident about).

Who pays: bookkeepers, small finance teams, agencies reconciling vendor bills. Two warnings. Scanned PDFs come out of Extract From File as garbage or nothing at all, so you'll need an OCR step in front for those, and clients never mention that half their vendors send scans. And that confidence flag at the end matters more than the extraction: a wrong number that lands silently in QuickBooks is worse than no automation.

7. Website or price change monitor (beginner). Watch a page, tell the client when it changes.

Node chain: Schedule Trigger (hourly/daily) -> HTTP Request (fetch the page) -> HTML Extract node (grab the target value) -> Compare against the last stored value (Data Store or a Sheet) -> IF (changed) -> Slack/Email node (notify).

Fastest thing on this list to build. Competitor pricing, job boards, stock levels, whatever. Clients love it because it feels like a superpower and you built it before lunch. Great first paid deliverable.

8. Scheduled data sync between two SaaS tools (intermediate). Two tools that don't talk, kept in agreement on a schedule. No more exporting a CSV every Monday.

Node chain: Schedule Trigger -> Source node (read records from tool A, e.g. Airtable or a database) -> transform/map fields -> Destination node (upsert into tool B, e.g. a CRM or Google Sheet) -> log the run.

Who pays: basically every business running two systems that "almost" integrate. Unglamorous, permanent, and the kind of thing that sits on a retainer for years because nobody wants to own it internally.

AI and reporting workflows

AI makes these possible. The boring scheduling and formatting around it is what makes them sellable.

9. Daily metrics digest (intermediate). Numbers from a few APIs every morning, formatted, in their inbox before coffee.

Node chain: Schedule Trigger (7am) -> multiple HTTP Request nodes (Stripe revenue, GA4 traffic, ad spend) -> Merge node -> AI/LLM node (write a short plain-English summary of what changed) -> Email or Slack node (send the digest).

Who pays: founders and marketing managers who currently log into five dashboards to answer "how did we do yesterday?" The AI commentary is what makes it feel premium instead of a data dump. Bundle it or sell it standalone.

10. Inbound email classify and auto-draft reply (advanced). Read the email, work out what it wants, draft a reply the client only has to approve.

Node chain: Gmail/Outlook Trigger -> AI/LLM node (classify intent + urgency) -> Switch (sales / support / spam / personal) -> AI/LLM node (draft a reply using a knowledge base or past emails) -> Gmail node (save as draft) -> Slack node (notify: "draft ready to review").

Who pays: solo founders and small teams buried in repetitive email. This is the high end of the list, $1000-2000/mo territory, and people pay it quietly and keep paying. Human approves before send. That rule is not negotiable, for quality and for trust.

Turning these into a business, not a hobby

Look at the pattern. Every workflow up there has a trigger a business cares about, a transformation that used to eat someone's afternoon, and a notification so a human stays in the loop. That structure is the product. The specific tools barely matter.

The builders making real money mostly aren't chasing the flashiest demo. They pick two or three of these, get properly good at them, and sell them as a monthly service. One person running five of these for five clients is somewhere in $2,500-10,000/mo without hiring anyone. That's the one-person-agency model, and automation is probably the cleanest way into it right now.

If you want the structured path instead of piecing it together from scattered videos, that's what CodingPhase is built for. The AI automation career path walks you from your first workflow to packaging and pricing client work, and how to become an AI automation specialist lays out the full roadmap.

Membership is $49/month, or $250/year (roughly $21/mo billed annually), with a 7-day money-back guarantee, so you can try it and bail if it isn't for you. If you'd rather pay once and own it, the Tech Accelerator is $1,500 lifetime and adds live weekly mentorship on top of every course. Either way you get the automation path plus the one-person-agency playbook that turns these workflows into invoices.

FAQ

Do I need to know how to code to build these n8n workflows? No, and that's the point. n8n is visual, so you drag nodes and connect them. Light JavaScript in a Code node helps for the advanced ones, mostly custom transforms and awkward data shapes, but you can build and sell everything beginner and intermediate on this list without writing much at all.

How much can I actually charge for one of these? It depends on the pain you're removing. A simple page monitor might be $150-300/mo. A support triage or email-drafting system that saves a team hours a day can run $1000-2000/mo. Price the value, not your build hours. The client is buying time back, and they don't care that you finished it on a Sunday.

Is n8n better than Zapier for client work? For agency and reseller work, usually yes, because self-hosting kills the per-task fees that wreck your margins at scale. Zapier is faster to start and has more prebuilt integrations, which sometimes matters more than the money. I break the tradeoffs down in n8n vs Zapier so you can pick per client instead of picking a side.

What's the fastest workflow to build first? The page or price monitor (number 7) or the lead-routing flow (number 1). Both beginner-level, both solve something a business feels immediately, and both make a first paid deliverable you can point at when you pitch the next client.

Won't AI hallucinations get me in trouble? Only if you let AI act without a human check. Every AI workflow here ends in a draft, a queue, or an approval step. Never a direct send or post. Keep a person on the trigger for anything customer-facing and you get the speed without betting the relationship on a wrong guess.

Pick one workflow from this list. Not five. Build it end to end, break it, fix it, and get it running reliably for one real use case, even if that use case is your own inbox. That one finished workflow teaches you more than ten half-built demos, and it's what you show the first client who asks what you can do.

You don't need to be an expert to start. You need one working automation and the nerve to charge for the next one. If you want a guide the whole way through, we've got you inside CodingPhase. Go build something that someone would miss when it stops.