AI-driven lead generation and personalised outreach at scale

The challenge

The client ran a niche B2B service and needed a repeatable outbound motion without a sales team. Manual prospecting was slow, generic templates got ignored, and most lead lists contained stale or invalid emails that burned domain reputation. The client also wanted to own the learning loop — not just run a campaign, but understand which offers, subject lines, and prospect segments actually convert.

What we built

We built a fully automated pipeline from prospect discovery to booked call. Google search scraping combined with the Google SERP API surfaces targeted prospect companies based on custom queries. The pipeline fetches each site's content to extract context for the email writer. GPT-4 reads that context and writes a personalised opening line and offer framing for each lead — at scale, without it reading like a template. Hunter.io enriches found contacts with verified email addresses and confidence scores, filtering out low-confidence records before they touch the sending domain. Verified leads go into Hunter.io sequences for structured follow-up. Every two weeks, we reviewed open rates, reply rates, and conversion by segment, then updated the offer framing, subject line variants, lead query strategy, and sequence timing based on real data.

The results

• Hundreds of targeted, personalised emails sent weekly with no manual writing
• High deliverability maintained through Hunter.io email verification before sequencing
• Bi-weekly optimisation loops improved reply rates each cycle based on live stats
• Offer copy, subject lines, and prospect targeting all evolved with the data — not set-and-forget
• Client retained sole ownership of the pipeline, data, and results with no third-party dependency