App Masters · 2026-06-06
Generative Engine Optimization (GEO): Grow Downloads With AI Search
Steve P. Young interviews David Andersen (co-founder/CTO of BrandViz.AI, ex-Google) on getting apps recommended by ChatGPT, Gemini and Perplexity. Covers BrandViz client cases (an Australian e-prescription app going from 5-10% to ~50% AI mention rate in 30 days; a meditation app's blogger outreach), a Pokemon card app with 11,000 pages, Momego's Q&A description and free website, app-blocker citation data (Aten, Freedom, AppBlock), and short audits of NutraMind and Aten.
Play the episode9 tactics
Numbers mentioned
| Who | Metric | Value | Context | When |
|---|---|---|---|---|
| marketConsumers overall | share | 58% approximate | Share of consumers who use AI tools to research products (discovery, shortlisting, comparison); varies by industry | 6:04 |
| benchmarkConversion of AI-chatbot referral traffic vs Google search traffic (studies cited by guest) | conversion rate | +4–5 x approximate | AI-referred visitors convert 4-5x as often as visitors from Google; lower volume | 52:13 |
| marketApps/businesses on average | share | 10% approximate | Share of conversions driven through AI on average; usually the second or third channel | 52:47 |
| aggregateSome app founders the guest spoke to who optimize for AI search | share | 50% approximate | Share of downloads attributed to ChatGPT; depends on the category | 7:44 |
| benchmarkNumber of options an AI chatbot presents in a recommendation | count | 3–5 options approximate | AI answers list 3-5 options; apps not on the list get no visibility | 8:38 |
| benchmarkHow long content keeps its AI-search visibility without updates | time to milestone | 3 months approximate | Content not updated for ~3 months sees a sharp drop-off in AI citations | 36:07 |
| benchmarkNumber of website content pieces for a baseline of AI-search visibility | count | 10–20 pages approximate | Pieces of content needed for an initial visibility push | 28:13 |
| Australian e-prescription app (BrandViz client) | share | 5–10% approximate | Share of AI answers across its space that mentioned the app, before the 30-day project (App Store listing only) | 40:54 |
| Australian e-prescription app (BrandViz client) | share | 50% approximate | Share of AI answers mentioning the app after ~30 days of fixes and content; its website became the most-cited domain | 41:27 |
| Australian e-prescription app (BrandViz client) | count | 20 content pieces/month approximate | Pieces of website content published over 30 days | 41:27 |
| Australian e-prescription app (BrandViz client) | count | 4–5 competitors approximate | Real competitors in its niche (less contested than most) | 42:06 |
| Pokemon card grading app | count | 11K pages approximate | Website pages, one per Pokemon card and variation | 26:39 |
| Pokemon card grading app | share | 50% approximate | Share of traffic coming through AI (mostly ChatGPT); 'more than 50%' | 52:47 |
| Aten | share | 10% approximate | Share of AI answers to comparison/alternative questions that mention the app by name (BrandViz test) | 18:41 |
| Aten | share | 40–50% approximate | Split of the app's AI citations between App Store listing and its website: first given as 50/50, later as roughly 40/60 | 39:21 |
| aggregateApps cited by AI chatbots for ~1,200 app-blocking category questions (BrandViz test) | count | 136 apps | Distinct apps among 315 App Store citations; AI reaches far down the long tail of apps | 38:46 |
| AppBlock | share | 70–80% approximate | Share of AI citations coming from its own website rather than the App Store (also true of other larger competitors) | 39:21 |
| Freedom | count | 320 queries approximate | AI answers mentioning Freedom out of ~600 app-blocker queries run by BrandViz | 1:03:57 |
| App Masters client fasting app | rank | 1 rank | Top app recommended by ChatGPT for fasting | 7:20 |
| Freedom | price | USD 40 | In-app purchase price seen on its App Store listing | 1:15:41 |
| Aten | price | USD 20 | Subscription price on its paywall (3-day trial); host thinks it is too cheap | 1:15:41 |
| AppBlock | price | USD 30 | In-app purchase price seen on its App Store listing | 1:16:05 |
| Opal | price | USD 90/year | Annual plan price seen on its App Store listing | 1:16:05 |
| NutraMind | count | 6 ratings | App Store ratings at time of audit | 1:11:20 |