n8n automation for B2B revenue teams

Per-task pricing is what puts most RevOps teams on this page. A dozen ordinary Zaps, form fills into HubSpot, Slack notifications when deals change stage, a push into a Google Sheet for reporting, nothing exotic, and the automation bill still lands in the same range as the CRM bill. Zapier charges for every action step in every run, so the cost of a multi-step workflow climbs with volume rather than with complexity.

Rebuild the same set in n8n and the pricing model changes shape. You pay per workflow execution, not per step, and n8n Cloud starts at $24 a month. For a stack of multi-step automations, that is the whole argument in one line.

That comparison comes up in nearly every conversation I have with a Series A or B company about their RevOps stack. Not because n8n is always the right answer, but because most teams are running expensive, shallow automation when they could be running cheap, deep automation. The distinction matters a lot at the scale where you're adding headcount and trying to keep revenue operations tight.

I've built n8n workflows for GTM teams across SaaS, professional services, and marketplace businesses. Here's what I've learned about where it actually earns its place in a B2B stack.

What n8n is (and what it isn't)

n8n is a workflow automation platform. You connect apps with nodes, define triggers, and automate sequences of actions without writing full code. It has about 1,000 native integrations, but more importantly, it has an HTTP Request node that makes virtually anything you can reach via API work as a step in a workflow.

The self-hosted version is open-source and can run on a $6/month VPS. n8n Cloud starts at $24/month and gives you a managed environment without infrastructure work. For most RevOps teams, Cloud makes more sense unless you have a developer who wants to own the server.

Where it differs from Zapier and Make:

• Pricing: n8n charges per workflow execution, not per task or operation. A 40-step workflow costs the same as a 4-step one. For complex B2B automations, this is a big deal.
• Complexity ceiling: n8n handles branching logic, loops, sub-workflows, error handling, and custom JavaScript nodes. Zapier handles most of this, but the interface gets unwieldy fast. Make handles it visually. n8n sits between them, closer to Make in capability but with a steeper initial learning curve.
• AI nodes: Since the January 2026 release, n8n ships with 70+ AI nodes including LangChain integration, persistent agent memory, and native support for self-hosted LLMs. If you want to build AI-powered GTM automations without stitching together five different services, n8n has become the most practical option.

What n8n isn't: it's not a data warehouse, a BI tool, or a replacement for your CRM. It's the plumbing between those systems.

Moving from Zapier to n8n: what changes on the bill, and how to migrate

The cost difference is a difference in what gets counted, so it is worth being precise rather than quoting a multiple.

Zapier counts a task "whenever Zapier successfully completes a unit of work for you". Triggers and polling do not count; every successful action step does. The Professional plan starts at 750 tasks a month for $29.99 billed monthly or $19.99 billed annually, and climbs through tiers to 2 million tasks (Zapier pricing). n8n counts "a single run of your entire workflow. It doesn't matter how many steps are in the workflow or how much data it processes." Cloud starts at EUR 20 a month for 2,500 executions, and Pro is EUR 50 for 10,000 (n8n pricing).

Put the five-step enrichment workflow from this page against both meters. On Zapier, one lead costs five tasks, so 100 leads a week is a little over 2,000 tasks a month and sits well above the entry tier. On n8n, one lead is one execution, so the same volume is about 430 executions a month and fits inside the entry tier with room for the other workflows on this page. The arithmetic runs the other way for a one-step Zap, which is why the recommendation further down still sends simple linear automations to Zapier.

Two things people get wrong when they switch:

• The migration is a rebuild. Zapier exports do not load into n8n, because the two are built differently. Rebuild each Zap from its description: a Zapier trigger becomes an n8n trigger node or a Webhook node, which accepts GET, POST, PUT, PATCH, DELETE and HEAD, supports basic, header or JWT auth, and has separate test and production URLs (n8n docs). Filters become IF nodes, formatters become Set or Code nodes. App connections get re-authorised in n8n; there is no copying credentials across.
• HubSpot allows one active webhook subscription per app. n8n's HubSpot Trigger covers created, deleted and property-changed events on contacts, companies, deals, tickets and conversations, but "if you activate a second trigger, the previous trigger stops working" (n8n docs). Route every HubSpot event through one trigger workflow that fans out, or use HubSpot's own webhook action in workflows, which pushes to an HTTPS endpoint and retries failed calls for up to three days on Data Hub Professional and above (HubSpot knowledge base).

The order that works: inventory every Zap with its trigger, apps, monthly task count and owner. Stand up n8n with a HubSpot credential, using a private app token scoped to only what the workflows write, of which HubSpot allows up to 20 per account (HubSpot developers). Rebuild the highest-volume Zaps first, because those are the ones paying for the migration. Run old and new in parallel for a week per batch with the new writes going to a test property. Then switch the Zap off.

The six n8n workflows worth building for B2B revenue teams

These are the automations I keep rebuilding. Not because they're clever, but because they solve problems revenue teams deal with every week. They are the revenue slice of a wider set. For the ones that run past RevOps into finance, onboarding, and content operations, I wrote those up as seven n8n use cases across the business.

1. Real-time lead enrichment and CRM write-back

The problem: a form fill comes in, it lands in HubSpot, and a rep gets a notification with a name and an email. No company size, no funding stage, no tech stack, no LinkedIn URL. The rep either researches manually (15 minutes) or calls blind (bad).

The n8n version:

• Trigger: HubSpot form submission (webhook) or new contact created
• Enrich: Call Clearbit, Apollo, or Clay API with the email or domain
• Score: Run ICP match logic (company size, industry, tech stack signals)
• Write back: Update HubSpot contact and company properties with enriched data
• Route: If ICP score is above threshold, create a deal and assign to rep. Below threshold, enroll in nurture sequence.
• Notify: Slack message to rep with enriched context

This replaces manual research and removes the 15-minute delay between form fill and qualified outreach. For teams with decent inbound volume (50+ leads/week), this is the workflow where the change in speed-to-lead is most visible, because the delay it removes was a human one.

2. Deal age and pipeline health alerts

Most HubSpot deal dashboards look fine until you look closely. There are deals with no logged activity for weeks still sitting in "Proposal Sent." There are deals where next steps are blank. There are high-value deals that have sat in the same stage for over a month with no meeting booked.

This workflow runs on a schedule (every weekday morning) and identifies stale deals:

• Trigger: Cron schedule (8 AM weekdays)
• Pull: HubSpot API call to get all open deals with last activity date
• Filter: Deals with no activity in 14+ days, or deals past average stage duration
• Notify: Slack DM to deal owner with deal name, stage, last activity, and value
• Escalate: If deal is stale by 21+ days and value is above threshold, ping the manager too

This doesn't nag reps constantly. It only fires when something genuinely needs attention. The visible effect is on the pipeline review: it gets shorter, because the obvious problems have already been solved before anyone sits down.

3. Job change monitoring for expansion and win-back

This one consistently surprises people. Your best buyers are the contacts who already bought from you. When they move to a new company, there's a window (usually 90 days) where they're still setting up their stack and receptive to solutions that worked for them before.

The workflow:

• Trigger: Cron (weekly)
• Pull: All closed-won contacts from HubSpot (or customers from your CRM)
• Check: LinkedIn API or Clay enrichment to check current job title and company
• Compare: Against the company and title in HubSpot record
• Flag: Contacts who have moved to new companies in the last 30 days
• Create: A new task or deal in HubSpot for the appropriate rep with context
• Notify: Slack message with the contact name, old company, new company, new title

This is the workflow that pays for itself fastest. The contacts are already fans of the product. They just need someone to notice the move and reach out while the new job is still new.

4. Intent signal to outbound sequence trigger

If you're using a buyer intent platform like 6sense, Bombora, or even LinkedIn Sales Navigator's activity signals, the data is only valuable if it triggers action fast. Most teams look at intent dashboards once a week. The signal is cold by then.

The n8n version:

• Trigger: Webhook from intent platform when an account crosses a scoring threshold
• Check: Is this account already in HubSpot as a deal or active conversation? (prevents double outreach)
• Enrich: Pull current contacts at the account and their roles
• Decide: If a champion-level contact exists, route to account owner. If not, create a prospecting task.
• Enroll: API call to your sequencing tool (Apollo, Salesloft, Outreach) to enroll the right contact in the right sequence
• Log: Create activity in HubSpot to record the signal and action taken

The key thing this solves is the gap between "we have the data" and "someone acted on it." Intent signals have a half-life. A company researching your category today is much more receptive than the same company four weeks from now. Automating the handoff from signal to sequence compresses that response time from days to minutes.

5. Win/loss interview and close reason capture

Win/loss data is one of the most underused assets in B2B companies. Everyone wants it. Almost nobody collects it systematically. The usual process is: rep fills out a form in HubSpot when they close or lose a deal, but field completion is inconsistent and the data is thin.

The n8n approach:

• Trigger: HubSpot deal stage change to "Closed Won" or "Closed Lost"
• Wait: 24-48 hours (deals close at odd hours, give the rep a moment)
• Send: Automated email or Slack message to the deal owner with a 3-question form (why won/lost, what the alternative was, what would have changed the outcome)
• Parse: If you want to get fancy, run responses through an AI node for theme extraction
• Write: Structured responses back to deal record in HubSpot as a custom property and note
• Aggregate: Weekly Slack summary to sales leadership with close reason distribution

This builds a dataset over time. After 90 days, you have enough data to run actual analysis on why deals close. That input shapes messaging, objection handling, and competitive positioning in a way that gut instinct never does.

6. AI-powered meeting prep brief

This one uses n8n's AI nodes. Before a discovery call, the rep needs context: what the company does, who they're meeting with, recent news, HubSpot history, open tasks. Finding all of that takes 15-20 minutes. Most reps skip it.

The workflow:

• Trigger: Google Calendar event created (via Calendar API) with a contact linked to HubSpot
• Pull: HubSpot contact and company data, deal history, last email content, open tasks
• Fetch: Company news from a news API (newsapi.org or similar) filtered to last 30 days
• Generate: AI node (GPT-4o or Claude) synthesizes everything into a 1-page brief: company summary, contact background, previous interactions, suggested discovery questions, competitive context
• Send: Email or Slack DM to the rep 2 hours before the meeting

Reps who use this show up better prepared and the meetings go further faster. It is roughly three hours of build time, which makes it one of the cheapest workflows on this list.

What the n8n setup actually looks like for a RevOps team

For a typical Series A company (30-80 employees, HubSpot as CRM, some combination of Clay, Apollo, Slack, and Google Workspace), here's how I'd structure the n8n deployment.

You want separate workflows for separate concerns. Don't build one giant workflow that does everything. When something breaks, you want to know exactly where.

Core workflows for RevOps:

1. Lead enrichment and routing (runs on webhook)
2. Pipeline health alerts (runs on cron)
3. Win/loss capture (runs on deal stage change)
4. Job change monitoring (runs weekly)
5. Meeting prep brief (runs on calendar event)

Each workflow should have:

• Error handling nodes that fire a Slack alert when something fails
• Logging to a Google Sheet or your data warehouse for debugging
• Clear naming conventions so you can tell what each workflow does without opening it

The total build time for all five, starting from scratch, is about 12-15 hours for someone comfortable with APIs and logic flows. If you're newer to this, budget 20-25 hours.

On pricing: n8n Cloud at the Starter tier ($24/month) gives you 2,500 workflow executions per month. For the five workflows above, a company with 100 leads per week and 20 meetings per week would use roughly 1,200-1,500 executions per month. Fits comfortably. If you scale up or add more automations, the Pro tier at $60/month gives you 10,000 executions.

Where n8n doesn't fit

I want to be honest about where I wouldn't use n8n.

If your team isn't technical at all and has no one who can read a JSON payload or understand an API response, start with Zapier or Make. The learning curve is real. n8n's interface is good but it's not point-and-click the way Zapier is for simple tasks.

If you're mostly doing simple linear automations ("when X happens, do Y"), Zapier's 8,000+ native integrations mean you might not need to write a single line of JavaScript. That's genuinely easier.

If reliability is critical and you don't want to own the infrastructure, be thoughtful about self-hosting. n8n Cloud solves this, but self-hosted n8n needs maintenance. Crashing automations during a campaign launch is painful.

And if your main need is marketing automation (email sequences, lead nurturing, behavioral triggers), HubSpot's native automation is probably sufficient. You don't need n8n for what HubSpot does natively.

The sweet spot for n8n is the stuff in between: multi-step flows, data transformation, API calls to tools HubSpot doesn't integrate with natively, AI-augmented workflows, and anything where Zapier's per-task pricing would get expensive at scale. The AI-augmented end of that list is its own discipline once a model is making decisions, and the mechanics of building, testing and monitoring an agent against HubSpot are worth reading before you give one write access. Which layer should do the coordinating once there is more than one agent is a separate decision, compared in our guide to AI agent orchestration for RevOps.

Measuring whether the automation is working: attribution and the KPIs to track

Readers ask which metrics show whether one automation setup attributes revenue more accurately than another. There is no single number, but there is a short set that separates a setup that is working from one that is producing dashboards.

Attribution accuracy, measured as agreement. Pick one HubSpot attribution model as the house view. HubSpot offers first interaction, last interaction, linear, U-shaped, W-shaped, time decay and full path, with contact-create reports on Marketing Hub Professional and deal and revenue attribution on Enterprise (HubSpot knowledge base). Then measure the share of closed-won deals where the model's first and last touch match what the rep recorded as the source. A setup that writes source properties automatically at the moment of enrichment scores higher here than one where reps fill them in later from memory, and the gap between the two is your attribution error.

Source and write coverage. Share of new contacts with original source, lifecycle stage and ICP score populated within an hour of creation. This is the metric the enrichment workflow moves, and it is the input to every attribution report downstream.

Speed to first touch. Median minutes from contact creation to first logged activity, read from HubSpot timestamps. Measure it before you build the enrichment workflow so the comparison is yours and not a vendor's.

Stale pipeline share. Share of open deals past their stage's median age. The health-alert workflow should push this down and keep it there.

Automated write errors. Count of workflow runs that hit the error branch, per workflow, per week. If this is not on a dashboard, the first sign of a broken integration is a rep complaining.

Suppression and deliverability, if you automate outreach. Two guardrails belong in the workflow rather than in a policy document. Suppression: with HubSpot's data privacy settings on, legal basis is a contact property and marketing email can be limited to contacts with consent (HubSpot knowledge base); the sequence-enrolment workflow reads that property and skips anyone without a basis, and a deletion in HubSpot triggers removal from the sequencer. Deliverability: the workflow verifies the email before enrolment and caps daily sends per sending inbox, so a bad enrichment run cannot burn a domain. Track bounce rate and spam complaint rate per sending domain weekly.

Comparing two setups is then straightforward: same month, same metrics, and the one with higher coverage, higher agreement and a lower error count is the one attributing more accurately. Everything else is opinion.

How this fits into a broader RevOps stack

n8n works best as the orchestration layer in a stack that already has the right data and the right tools. If your CRM data is garbage, automating workflows on top of it will amplify the garbage. Start with data quality.

The stack I usually recommend for Series A/B B2B companies:

• HubSpot as the CRM and marketing automation hub
• Clay for list building and enrichment
• Apollo or Outreach for sequencing
• n8n as the connective tissue between all of them

n8n handles the tasks that fall between these tools: writing Clay enrichment data back to HubSpot, triggering sequences from HubSpot events, building custom deal routing that HubSpot's built-in routing can't do, and sending the right notifications to the right people at the right time.

That stack is the execution layer of something larger. We set out the AI GTM strategy these workflows serve separately: the planning phase that decides what to automate, the intent signals worth wiring up, and the Claude, HubSpot and n8n workflows that run them.

This is the shape of our AI automation work, and the outcomes typically show up in two areas: faster lead response and cleaner pipeline data. Both of those eventually show up in win rate.

For companies that want to build this themselves, the n8n documentation is genuinely good. The community templates library has solid starting points for HubSpot and CRM workflows. Build the enrichment workflow first. It delivers the most visible impact in the least time.

If you want help building or auditing your current automation stack, we do that as part of our CRM and RevOps work. Happy to do a fast assessment of what's worth automating and what's already covered by your existing tools. Reach out here.

FAQ

What is n8n automation and how does it work?

n8n is a workflow automation platform that connects apps and services through a visual interface. You define triggers (something that starts the workflow), actions (what happens next), and conditions (branching logic). When the trigger fires, n8n runs the sequence automatically. You can connect HubSpot, Slack, Google Sheets, Clay, Apollo, and essentially any tool with an API. The self-hosted version is free; n8n Cloud starts at $24/month.

How is n8n different from Zapier for B2B teams?

The biggest difference is pricing and complexity ceiling. Zapier charges per task (each action in a workflow costs a task), so a 10-step workflow costs 10x more per execution than a 1-step one. n8n charges per workflow execution regardless of steps. For complex RevOps automations with 10-30 steps, n8n is usually 5-10x cheaper. n8n also handles JavaScript nodes, loops, sub-workflows, and AI agents that Zapier handles less naturally.

Which n8n workflow should a B2B team build first?

The lead enrichment and routing workflow. It fires the moment a new contact enters your CRM, enriches the record with company data and ICP scoring, and routes high-fit leads to reps immediately. For most teams, it has the fastest and most visible impact on pipeline because it directly improves speed-to-lead and data quality at the same time.

Can n8n connect to HubSpot natively?

Yes. n8n has a native HubSpot node that supports reading and writing contacts, companies, deals, tasks, and engagements. You authenticate once with your HubSpot API key or OAuth and all HubSpot operations are available as workflow steps. You can also use webhooks to trigger n8n workflows from HubSpot events in real time.

Do I need a developer to use n8n?

Not necessarily, but it helps to be comfortable with APIs and JSON. Simple workflows can be built without any code using n8n's visual interface. More advanced workflows (custom scoring logic, data transformation, AI integration) typically require writing some JavaScript in n8n's Code node. For RevOps teams with a technical operator or a RevOps engineer, n8n is very accessible. For fully non-technical teams, Zapier or Make will have a shorter ramp-up time.