Part of the management buyout of Disciple (from bankruptcy), then drove it to net profitability.
- Remade the inbound engine end-to-end — audited the Google Ads account and found the algorithm training on raw call volume not any meaningful signal. Re-instrumented the stack (Google Tag Manager custom conversion events based on form response) so it optimised on qualified intent, and then rewired agency feedback loop with sales around CRM outcomes. Trained sales on how to properly categorise leads. Cut spend from $60k to $15k/month while holding qualified lead volume flat — CAC down 67%.
- Built the outbound motion from the phones up — took the calls myself, wrote the playbooks off live conversations, then hired and ran an SDR team across the UK, US, Canada, Australia & NZ. Modelled it as a separate channel with its own economics — targeting ~$60k deals vs the $12k inbound average — closing $120k in revenue. We ultimately pulled the channel to focus purely on inbound.
- Vibecoded the full outbound pipeline (pre–Claude Code) — shipped a working v1 solo over a weekend: scraping, enrichment, LLM lead-scoring. Proved the channel, then industrialised it with engineering on Airflow. Latest iteration matches prospects to campaign angles via vector embeddings on their own website language — targeting by resonance, not firmographics.
- Led the JTBD research that repositioned Disciple as the mobile-first community platform — read better-funded rivals’ PLG incentives from their hiring and CEO narratives, took the sales-led segment they weren’t prioritising, and killed the web plan that was pulling us toward the wrong buyer, enabling the move upmarket.