Of the tools I install for B2B SMB clients in 2026, n8n is the one that produces the most leverage per dollar. The platform sits between the marketing stack (HubSpot, GA4, Search Console, CRM) and the AI APIs (Anthropic, OpenAI, Google), and runs the connective workflows that an agency would otherwise bill out as ad-hoc operations work. The math is straightforward: a self-hosted n8n instance at CAD $10 to $30 per month, plus 20 to 40 hours of one-time build, replaces roughly CAD $1,500 to $3,500 per month of agency operations time.

June 2026 pricing update: n8n removed the active workflow limit from all plans earlier in 2026, then shifted to execution-based billing — Starter at roughly €24/month (2,500 executions), Pro at €60/month (10,000 executions), Business at €800/month (40,000 executions), with annual billing 17% cheaper. Self-hosted Community Edition remains free with unlimited executions. For SMBs running fewer than 2,500 executions per month, the Starter cloud tier covers the six workflows in this post with room to spare. For teams with a developer, self-hosting on Hetzner, Railway, or Render at CAD $10 to $30 per month is still the most cost-effective option.

This post is the six workflows I build first for a new SMB engagement. The order matters — each workflow assumes the previous ones are in place — and the total build time is 12 to 25 hours for an operator who already knows the platform. For SMBs new to n8n, add a 10 to 20 hour learning curve before starting.

Workflow 1: Lead enrichment from form submissions

The first workflow because the value is immediate and the build is simple. The shape:

  1. Webhook trigger from the contact form (HubSpot form, Webflow form, or a custom endpoint).
  2. Apollo or Clay API call to enrich the submitted email with company data: industry, employee count, revenue band, recent triggers.
  3. Score the lead against the ICP attributes documented in the ICP playbook.
  4. Write the enriched record and score back to HubSpot.
  5. Notify the sales team in Slack with the enriched profile and the ICP score.

The total node count is typically 8 to 12 nodes. Build time: 2 to 4 hours. The first-month value is usually a 30 to 60% reduction in lead-qualification time, plus a noticeable improvement in sales-marketing alignment because both teams are now looking at the same enriched record.

Workflow 2: Daily GSC and GA4 digest

The second workflow because it builds the data-review habit that everything downstream depends on. The shape:

  1. Cron trigger at a consistent time each morning.
  2. Google Search Console API call: yesterday's clicks, impressions, top queries, top pages.
  3. GA4 Data API call: yesterday's sessions by channel, conversions, AI-attributed sessions from the custom channel group described in the AI attribution post.
  4. Format into a Slack-ready summary or an email body.
  5. Send to the marketing channel.

Build time: 3 to 5 hours, mostly spent on the API authentication setup the first time. Once the credentials are in place, the workflow runs autonomously. The behavioural change is the real value — operators who get a daily digest into Slack start noticing trends within two weeks that they would not have noticed in a monthly report.

Workflow 3: Weekly content briefing

The third workflow generates the content brief from the data. The shape:

  1. Cron trigger weekly (Monday morning works well).
  2. GSC API call: queries from the last 30 days with high impressions and low CTR.
  3. AI API call (Claude or ChatGPT) to draft a content brief: target query, working title, 3–5 H2 outline, FAQ questions, internal link targets, primary entity references.
  4. Post the brief to a content channel or write it to a Notion/Trello content queue.

Build time: 4 to 6 hours. The brief is not publishable as-is — a senior operator reviews and shapes it before a draft is written. The leverage is that the brief is automatically grounded in real GSC data, so the content roadmap stops being a guess and starts being a response to demonstrated demand.

Workflow 4: AI Overviews citation tracker

The fourth workflow operationalizes the citation tracking covered in the AI Overviews citation tracking post. The shape:

  1. Cron trigger weekly.
  2. For each of the 20–50 priority queries, call a SERP API (SE Ranking, DataForSEO, or similar) with AI Overviews enabled.
  3. Parse the response for citations of the target domain.
  4. Write the results to a Google Sheet or database.
  5. Send a weekly summary of citation changes (new citations earned, citations lost) to Slack.

Build time: 4 to 6 hours. The SERP API cost is the limiting factor — running 50 queries weekly through DataForSEO costs roughly USD $30 to $60 per month, which is the right price for the visibility it produces.

Workflow 5: Schema validation and refresh trigger

The fifth workflow catches the regressions that quietly cost rankings. The shape:

  1. Cron trigger weekly.
  2. For each page in the high-priority page list, fetch the HTML and extract the JSON-LD schema blocks.
  3. Validate the schema against expected types (Article, FAQPage, Service, Organization).
  4. Flag any pages where schema is missing, malformed, or out of date.
  5. Notify the operator with the list of pages needing refresh.

Build time: 3 to 5 hours. The value compounds over time — sites that catch schema regressions within a week instead of a quarter avoid the silent ranking drift that affects most SMB sites in the 12 months after launch.

Workflow 6: Monthly client (or stakeholder) report

The sixth workflow turns the daily and weekly data into a monthly artifact a non-technical stakeholder can read. The shape:

  1. Cron trigger on the first business day of the month.
  2. Pull aggregated data from GSC, GA4, the citation tracker sheet, and HubSpot.
  3. Run the data through an AI summarization step that produces a short executive narrative (3 to 5 paragraphs) plus a structured metrics table.
  4. Generate a Google Doc or PDF with the report.
  5. Send to the stakeholder distribution list.

Build time: 4 to 6 hours. The workflow replaces the most time-consuming recurring operations task in most SMB marketing operations: the monthly report. Operators who build this workflow recover 4 to 8 hours per month and produce a more consistent report than the manually assembled version.

What I deliberately do not automate

Three things I leave manual on purpose.

Content publishing decisions. The AI draft is automated; the decision to publish, to revise, or to kill is not. The senior operator's judgement on whether a piece of content represents the business well is the most important checkpoint in the system, and removing it produces median-competitor output efficiently.

ICP changes. The ICP is reviewed quarterly by the operator with input from the sales team. Letting an automation update the ICP definition without human review is a category of mistake that compounds.

Customer interview synthesis. The interviews are the source of strategic insight; the synthesis is too important to delegate. AI assists with transcription cleanup and pattern surfacing, but the synthesis is operator work.

Six workflows, one afternoon of build time, months of recovered agency fees

n8n is not a glamorous tool. It is the workflow runner that quietly does the connective work an agency would otherwise bill for. The six workflows above, built in order, produce a marketing operations foundation that scales from 5 to 50 person businesses on a budget that is a fraction of agency operations cost. For the broader picture of how this fits into the AI marketing stack, the best AI marketing tools post covers the full tool list, and the services overview covers what gets owned and reviewed each month.

The three n8n workflows every B2B SMB should build first

If the full list of six workflows feels like too much to start with, begin here. These three deliver the most value for the smallest build investment and create the data foundation the other three depend on.

First: the lead enrichment workflow. Every new form submission or inbound lead gets automatically enriched with company size, industry, and LinkedIn data before it lands in your CRM. This takes a task that a sales person used to do manually — often hours after the lead came in — and makes it instant. The workflow pays for itself the first week a fast-moving lead converts because the sales rep already knew their context.

Second: the weekly content performance digest. GSC clicks, GA4 sessions, and conversion events pulled into a single Slack message every Monday morning. No login required, no report to build. The person responsible for content knows what is working without spending 45 minutes in analytics dashboards.

Third: the AI content brief generator. A keyword or topic goes in; a structured content brief with target audience, key questions to answer, competitor gaps, and suggested schema markup comes out. Build this before any other AI-assisted content workflow because it improves every piece of content that follows it.

Build these three first, run them for a month, and then layer in the remaining three. The order matters because each workflow teaches the operator what the system can do — which makes the later builds faster and better scoped.