Most brands have no idea whether AI recommends them. This is what the first day of fixing that actually looks like, with real numbers and nothing hidden.
Paris Childress of Hop AI and Simon Young of Question.Marketing take one of Simon's clients, WhatsApp Business Platform Stitch AI, and start building its visibility inside ChatGPT, Google AI Overviews and Google AI Mode in public. No case study gloss. Day one, live, including the baseline nobody wants to publish: Stitch is mentioned in one prompt out of thirty.
Over the next 60 days this series follows the whole build. This first session covers how the prompt set gets chosen, how share of voice is measured and why it is harder than SEO ever was, how a knowledge base of 1,335 sales call recordings gets turned into content AI models will cite, and what "information gain" actually means when everyone is drowning in AI slop.
WHAT YOU'LL TAKE AWAY
How to attribute leads from AI when there is no click to track
Why the prompt set is the part most businesses get wrong
How to find the words your customers really use, using competitor reviews
Why first-party knowledge beats volume when you want to be cited
How to baseline your share of voice before you spend a penny on content
CHAPTERS
00:00 Who's on this and what we're building
01:00 From black hat SEO to answer engines: Simon's route in
02:00 Where did all the clicks go?
03:00 Calling out AI referral numbers nobody can prove
04:00 Is GEO a real channel, or SEO with a new sticker?
05:00 Zero-click search: why an AI referral leaves no trail
07:00 The "how did you hear about us?" fix for AI attribution
08:00 What Stitch AI does, and why the category is hard to name
10:00 "Bulk WhatsApp": the term customers search, not the one you'd pick
13:00 Using G2 to find the category your competitors sit in
14:00 Mining 500 competitor reviews for the voice of the customer
15:00 Inside the knowledge base: 1,335 sales calls in 24 hours
17:00 Information gain explained: scoring what models haven't seen
18:00 AI slop, the Anthropic watermark and where LinkedIn lands
21:00 Does AI-generated content get penalised? The honest answer
23:00 Building 30 prompts across top, middle and bottom of funnel
25:00 Can you retire a prompt once you own it?
26:00 The baseline: mentioned in 1 prompt out of 30
27:00 Which engines we measure, and which we ignore for now
29:00 Turning the gaps into a content plan
31:00 Case studies, comparison pages and keeping data safe
32:00 The 60-day goal and what happens on Day 2
ABOUT THE GUESTS
Paris Childress is the founder of Hop AI and GEOForge, the platform used throughout this session to build the knowledge base, track prompts and measure share of voice.
https://hoponline.ai
Simon Young runs QuestionMarketing, a done-for-you AI implementation and answer engine optimisation consultancy in Doncaster, South Yorkshire. He has been publishing on answer engine optimisation since October 2019, more than three years before ChatGPT existed.
https://question.marketing
https://www.linkedin.com/in/aeo
Stitch AI is an official Meta business partner and WhatsApp communication platform for businesses running messaging at scale.
SUBSCRIBE for Day 2, where the first content drafts come out of the knowledge base and we see whether the share of voice moves.
Questions for the next session go in the comments. We read them live.
Generative engine optimization (GEO), also called answer engine optimisation (AEO) or AI search optimisation, is the practice of getting a brand named and cited inside AI-generated answers rather than ranked in a list of links. This session covers AI visibility, share of voice measurement, LLM citations, prompt tracking, ChatGPT, Google AI Overviews, Google AI Mode, zero-click search, information gain, first-party knowledge bases and build-in-public GEO strategy for startups and B2B SaaS.
#GEO #AEO #AIVisibility #GenerativeEngineOptimization #AISearch #ChatGPT #B2BSaaS