Marketers and small business owners rely heavily on LinkedIn to find potential customers and initiate conversations. There are hundreds of activities recorded from all the connections that a person has and some of them could be valuable signals leading to sales opportunities. However, it is not possible to keep a track of all the updates from connections on LinkedIn 24x7. That's why we've created this automation workflow using n8n and Google Gemini which helps you execute this task 24x7, automatically so that you don't miss any sales opportunities within your network. Here's the step-by-step guide: Step 1 - Daily trigger - Node: Schedule Trigger, run once at end-of-day. - Input: list of LinkedIn connection IDs (store as credential/variable). --------------- Step 2 - Fetch connection activity - Node: HTTP Request to LinkedIn API / provider. - Pull each connection’s recent updates for the day: posts, comments, likes, shared articles, job changes using LinkedIn's official APIs / SerpApi / ScrapingDog's LinkedIn Scraper API. - Use SplitInBatches to iterate safely and respect rate limits. --------------- Step 3 - Normalize activity items - Node: Set / Function. - Produce structured records: connection_name, profile_url, item_type (post/comment/like), text_snippet, post_url, timestamp, engagement_count. --------------- Step 4 - Summarize daily activity (Gemini) - Node: HTTP Request → Google Gemini. - Prompt: ask Gemini to turn a connection’s activity items into a 1–2 line summary (what they posted, tone, any request/need hinted). - Save Gemini summary per connection. Short example prompt to paste into Gemini: "Summarize these LinkedIn activity items for [name]. Output one short sentence describing what they’re up to today and whether they appear to be seeking vendors, hiring, or sharing product feedback." --------------- Step 5 - Enrich & match to your offering (optional) - Node: HTTP Request to Cura8.ai by navan.ai, Clearbit or Apollo. - Get company industry, size, recent news. - Combine company facts with the activity summary for context.

When I first started with n8n, I learnt that the easiest way to build complex workflows is by starting with tiny experiments. One of those experiments: 👉 Take an email list, automatically find the LinkedIn profiles, and update the sheet. The workflow: - Pull emails from Google Sheets - Extract name & company - Search for their LinkedIn profile - Write the link back into the sheet Simple, but powerful. 🙌 What excites me is how these little automations stack up. This one can plug into bigger systems for enrichment, scoring, and CRM updates. Start small → scale big. 🚀

For certain businesses, new opportunities open up when there's a change in the management or the target lead gets promoted. To find such prospects manually is nearly impossible for any marketing team. Hence, we've created this guide which will help you automate this process using n8n and Generative AI: Step 1 — Define ICP (user input) - Node: Manual Trigger or Form node to collect: target job titles/designations, industries, country/state/city, keyword(s) (e.g., “automation”), seniority level, company size range. - Save ICP as variables for downstream queries. ------------ Step 2 — Search LinkedIn for matches - Node: HTTP Request to LinkedIn API / Sales Navigator / SerpApi People Search. - Query using ICP variables: title keywords OR exact titles, industry, location, company size filters, and keyword in headline/about. - Use pagination + SplitInBatches for rate limits. ------------ Step 3 — Grab promotion signals (profile + posts) - For each candidate, collect available signals: current position start date, previous role timestamps, recent LinkedIn posts mentioning “promoted”, “excited to take on”, “new role”, or job change activity. - If using Sales Navigator/People API, pull position history and activity feed snippets. If using SerpApi, capture profile headline + recent post text. ------------ Step 4 — Use Gemini to classify “promoted in last 3 months” - Node: HTTP Request → Google Gemini API. - Ask Gemini to read profile snippet + position start dates + recent post excerpts and return JSON: {promoted:true/false, promotion_date, confidence, reason_short}. - Threshold: keep only promoted=true AND confidence ≥ 75 (adjustable). Suggested short Gemini prompt (single-line style you can paste): "Classify whether this LinkedIn profile or recent post indicates the person was promoted within the last 3 months. Return JSON: {promoted:true/false, promotion_date:YYYY-MM-DD or null, confidence:0-100, reason}." ------------ Step 5 — Enrich shortlisted prospects (optional)

There are hundreds of virtual events happening on LinkedIn every day. Did you know that a lot of these events allow you to see the other attendees once you RSVP for those events? That's where your potential customers can be found. Curious to know how and want try it for your business? Follow this guide: 1. Trigger - Node: Manual Trigger (so you can run it after RSVPing “Yes” to a webinar). - Input: LinkedIn event URL or event name (e.g., “Automation”). 2. Collect Webinar Attendee Data - Source: LinkedIn event “Networking” tab (accessible through LinkedIn API or a service like SerpApi). - Node: HTTP Request. - Output: list of attendee profiles (name, title, company, LinkedIn URL). - Note: respect LinkedIn’s API policies - don’t scrape unapproved. 3. Structure Attendee Data - Node: Set. - Extract fields: first_name, last_name, title, company, profile_url. - Store them in structured format for later enrichment and message generation. 4. Enrich Profiles (optional) - Node: HTTP Request to Cura8.ai, Clearbit, Apollo, or ZoomInfo. - Add company info: size, industry, funding, headquarters. - This context will improve personalization. 5. Create Hyper-Personalized Messages (GenAI) - Node: HTTP Request → Google Gemini API. - Prompt idea: “You are a marketing professional preparing to reach out to webinar attendees. Write a short, friendly LinkedIn message (3–4 lines) to {{first_name}}, {{title}} at {{company}}, who attended the {{event_name}} webinar. Acknowledge the webinar, show curiosity about their business, and suggest exploring if our offerings could help. Do not pitch aggressively. Output JSON: {message}.” - Gemini returns the personalized message text for each attendee. 6. Log Results - Node: Google Sheets. - Create a sheet with columns: name, title, company, LinkedIn URL, message_text, event_name, date_added. - Append each attendee with their personalized message. 7. Manual Review & Outreach - The marketing professional reviews the Google Sheet. - Copy/paste the hyper-personalized messages into LinkedIn manually (keeps it authentic and avoids automation risks on LinkedIn).

Most LinkedIn outreach fails because it feels irrelevant. Buyers today don’t want another templated pitch. - 61% of B2B buyers prefer a rep-free buying experience (Gartner). - 73% avoid suppliers who send irrelevant outreach. - Cold emails get ~5% replies vs LinkedIn DMs at 10%+ (Expandi, 2025). The signal is clear: prospects want context-aware, value-driven conversations. That’s exactly what AI agents deliver. They: ✅ Scan profiles & posts → highlight what really matters ✅ Analyze context → spot job changes, funding news, or pain points ✅ Draft relevant, timely messages & follow-ups Even LinkedIn itself is investing in AI: its Account Prioritization Engine boosted renewal bookings by 8% in internal tests. This isn’t about replacing humans. It’s about freeing you from repetitive research so you can focus on the conversations that matter. I’m currently building a LinkedIn lead gen agent that does this end-to-end - monitoring activity, detecting intent, and generating outreach messages you can actually use. 👉 What’s your biggest challenge in LinkedIn outreach - finding prospects, personalizing, or following up?