Webinar recap: Scale personalized LinkedIn outreach with AI

Live session · March 2026

Brian walked through a four-step workflow that combines lead search, profile enrichment, AI qualification, and AI-written messages, so your outreach feels personal even when it's running automatically.

Key takeaways

Pair dynamic filters with Watcher Mode for a "set and forget" lead flow 
Use signals like "currently hiring" or "posted on LinkedIn in the last 30 days" in your search. These filters refresh on their own, and Watcher Mode means PhantomBuster only pulls new leads each run, so your pipeline stays topped up without duplicates.

Enrich first, then qualify 
Before running AI classification, pull richer profile data with the LinkedIn Profile Scraper (headline, experience, job title). The more context your AI has, the better it can segment your leads.

Use AI to qualify, not just to write messages 
Run a simple classification prompt first, for example: “is this person a decision maker?” and filter your list down before writing a single message. Targeting the right people matters more than personalizing at volume.

The first message isn't a pitch, it's a soft hello 
Brian's approach: reference something specific on the person's profile, keep it warm and brief, and don't mention your product. The goal is just to make a connection before any selling starts.

Skip the connection note 
Research (and Brian's experience) backs this: you get roughly twice the acceptance rate without an invitation message vs. with one. People can spot a sales pitch before they even accept.

Ramp up gradually, ebb and flow, not spikes 
Start at ~5 connection requests a day, increase by ~20% every few days, and don't treat LinkedIn's rate limits as a target. Consistent, human-like activity patterns matter more than the specific number.

Q&A themes

What if I don't have Sales Navigator? You can still run this workflow using standard LinkedIn searches, Sales Navigator just helps you filter more precisely upfront.

How do you avoid sounding like a bot? Keep messages short, reference something specific from their profile, and skip the pitch in the first message. The goal is a conversation, not a conversion.

Do you need to enrich before you qualify? Not always, it depends on your lead source. A Sales Navigator search export includes enough data to qualify directly. A follower collector or liker scraper only gives you name + URL, so you'll want to enrich first.

Can this scale across a team? Search exports and profile scrapers support multiple LinkedIn accounts. But each outreach phantom is tied to one LinkedIn identity, so for team-wide outreach, set up a separate LinkedIn Outreach phantom per person.

▶️ Watch the full replay: 
Replay link

📄 Playbook: 
Craft personalized AI outreach to ICPs

💬 Let's discuss What part of personalizing outreach at scale feels hardest right now: finding the right signal, writing the prompt, or reviewing at volume?