In this episode of The Pump and Funnel Marketing Show, Araminta Robertson shares the GPT article framework her agency developed to help B2B brands increase visibility across ChatGPT, Gemini, Perplexity, and Claude. She reveals how clients now receive 30-40% of inbound leads via LLMs, establishing this as a major acquisition channel.

The methodology is concrete: select bottom-of-funnel topics, create 5-10 prompts per topic, then write one 500-1,000 word article per prompt. Araminta walks through the Fiska case study (visibility increased from 5% to 34% in under a month). She also explains why self-attribution ("How did you hear about us?") is the most reliable tracking method, since LLM-driven leads often appear as direct or organic in analytics.

CHAPTERS

[00:00] - LLM visibility as a major acquisition channel

[00:44] - Where LLM visibility fits in the Pump and Funnel model

[01:30] - What is LLM visibility and which LLMs matter most

[02:41] - The GPT article framework explained

[03:49] - How to pick bottom-of-funnel topics and prompts

[05:00] - Why publish on your own website, not LinkedIn or Medium

[07:46] - Best practices for 500-1,000 word GPT articles

[09:26] - Case study: Fiska's visibility increased from 5% to 34%

[10:16] - How to track LLM-driven leads with self-attribution

[11:58] - Why you still need SEO alongside GPT articles

[13:10] - Which industries benefit most from LLM visibility

[14:00] - How future-proof is the GPT article framework

[15:44] - Testing new approaches immediately rather than procrastinating

[17:11] - Why hosting in-person events speeds up marketing success

[19:36] - Secret software: Peek LLM visibility tracking tool

[20:14] - Next on the list: creating simple business videos for YouTube

[22:06] - Magical marketers: April Dunford and Louise Linehan

Find out more and download your free copy of the Pump and Funnel Quick Start Guide at https://pumpandfunnel.com.

In this episode of the Pump and Funnel Marketing Show, David speaks with Garrett Sussman, Director of Marketing at iPullRank, about hyper-personalisation in AI search. Garrett explains how AI search engines now generate unique answers for each user based on implicit context (Gmail, search history, location) and explicit context (specific query details), making traditional keyword targeting obsolete.

Garrett shares real examples: Google Personal Intelligence pulling hoodie purchase data from Gmail to make recommendations, and a study showing LLMs recommend different banks based on persona attributes. His advice: create persona-specific content, test using "search twins" (Gmail accounts populated with ideal customer interests), and track trends over time rather than chasing deterministic rankings.

CHAPTERS

[00:00] - Marketing to an audience of one

[02:01] - What is hyper-personalisation?

[05:48] - Why generic keywords are finished

[07:08] - LLMs already recommend based on personas

[10:43] - How Google Personal Intelligence works

[13:39] - Real test: AI reading Gmail to recommend hoodies

[16:05] - Measuring AI visibility (it's messy)

[18:03] - What Garrett wishes he'd done differently

[20:18] - Whisper Flow for AI-assisted dictation

[22:50] - Marketers to follow: Amanda Natividad and Angela Skane

[24:11] - Create "search twins" to test AI search results

Find more episodes and resources at https://pumpandfunnel.com.