All work
Insurance brokerageBG4 weeks build · ongoing optimization on retainer
SDR close rate from 4% to 11% after AI scoring.
LinkedIn lead-gen pipeline with AI ICP scoring and personalized first-touch drafting for an insurance broker.
4% → 11%
SDR close rate
5x
qualified meetings per week
8%
of leads reach the human
0
generic first messages sent
Challenge
The state we found.
A two-person sales team was running generic outbound on LinkedIn. The first-touch message was the same for everyone. Reply rates were low, qualified meetings were rare, and most calls went to people who would never buy commercial insurance.
Solution
What we built.
We built a pipeline that scrapes targets from LinkedIn Sales Navigator, enriches them with public company data, runs an AI ICP score, and drafts a personalized first message based on the prospect's recent activity. Only the top 8% reach the SDR's inbox.
Architecture
// 6 steps · production
- 01Sales Navigator search export → PhantomBuster scraping
- 02Enrichment with company size, industry, recent funding, and tech stack
- 03OpenAI ICP scoring (1–10) against the broker's historical close patterns
- 04Auto-drafted first-touch message referencing recent prospect activity
- 05Approved drafts pushed to LinkedIn via PhantomBuster scheduling
- 06Reply detection → auto-sync to GoHighLevel with full conversation history
Stack
PhantomBusterOpenAIGoHighLevelGoogle SheetsApify
Next case study
Read nextFour stores, one operations app — used daily from week one.
Final word
Book the call. Bring the bottleneck.
30 minutes. No deck. We either see the automation in your process — or we tell you it is not worth building.