GTM Sauce

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 episode

9 tactics

An e-prescription manager app · B2C · BrandViz client

Mentioned in ~50% of AI answers, up from 5-10%; became the most-cited site in its niche

Starting point: App Store listing gave a 5-10% AI mention rate; the website was fully JavaScript-rendered (AI crawlers do not run JavaScript) and not in search indexes. Over ~30 days: restructured the site so content is in the served HTML (server-side rendering), got it listed in search indexes (Google Search Console, Bing Webmaster, which feeds ChatGPT), then published ~20 pieces of content. Content answered problem-recognition questions users ask AI ('I deleted my SMS, I don't have my script, what do I do?') and ended with the fix (use our free app, it syncs everything automatically), plus pieces owning the rural/offline-access angle (which app to use when connection is spotty). Test: paste your URL into ChatGPT and ask what it can tell you about the company.

Workedseocontent
App Masters · 2026-06-06

A meditation app · B2C · BrandViz client

Not every blogger replied; no visibility numbers given

In a crowded category (meditation, dominated by Calm and Headspace, with problem questions answered by doctors), listed every blogger who showed up in AI answers for the space and reached out to them for guest posts or collaborations.

Unknownpartnershipsseo
App Masters · 2026-06-06

A Pokemon card grading app · B2C · founder the guest knows

Over 50% of traffic now comes from AI, mostly ChatGPT

Users mostly ask 'I have this card, what's the rating/value?'. Created a web page for every Pokemon card and every variation (~11,000 pages) with rough information about the card and a push to the app's AI photo grading to judge its quality.

Workedseoweb-to-app
App Masters · 2026-06-06

Wrote the App Store description as questions and answers users ask AI: 'What is Momego? Does Momego work in Sydney? What about Melbourne? Can I use it in Brisbane?'

Unknownasoseo
App Masters · 2026-06-06

Momego · B2C · live bus and train times app

Lots of traffic, some flowing on to the app; no numbers here

Built a website giving away a free layer of information (schedules, traffic times) with a funnel down to the app.

Workedweb-to-appseo
App Masters · 2026-06-06

Wrote a 'top couples apps' article: her own app at #1, competitors at #2-#4, with pros and cons of every app including her own. Guest adds: be charitable to competitors and frame it as which app fits which user.

Unknowncontentseo
App Masters · 2026-06-06

Aten · B2C · app blocker for focus

Later showed up as a source in AI answers about the app

Posted a public Reddit thread announcing the app launch ('this is the app we're launching').

Workedcommunitypr-launch
App Masters · 2026-06-06

Practitioner pattern · David Andersen (Co-founder of BrandViz.AI, which sells AI-search visibility tracking and optimization)

Citations drop sharply after ~3 months without updates; refreshes bring traffic bumps

AI favours recent content: citations fall off about 3 months after publishing. Do a light refresh of older pages (some rewording, a few added links, update for anything new) to get a bump in AI traffic to those pages. Also watch for old negative reviews and threads AI keeps surfacing, and respond on Reddit/community forums.

Workedseocontent
App Masters · 2026-06-06

Practitioner pattern · David Andersen (Co-founder of BrandViz.AI, which sells AI-search visibility tracking and optimization)

Gives a baseline of AI visibility in a month or two, per the guest; no numbers

1) Measure a baseline: how often AI names you for questions in your space. 2) Make the site readable: paste your URL into ChatGPT and ask what it can find; if it is JavaScript-only (common with Lovable/Claude Code sites), switch to server-side rendering; don't block AI bots; submit to Google Search Console and Bing Webmaster. 3) Publish 10-20 pieces: why you vs a bigger competitor, a landscape of the apps in the space and who each is for (conditional 'if you're X, use Y'), guides for the exact pain points that trigger a search, definitions of the category terms. 4) Ask ChatGPT for list sites and directories covering your space and get listed on them.

Workedseocontent
App Masters · 2026-06-06

Numbers mentioned

WhoMetricValueContextWhen
marketConsumers overallshare58%
approximate
Share of consumers who use AI tools to research products (discovery, shortlisting, comparison); varies by industry6:04
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 volume52:13
marketApps/businesses on averageshare10%
approximate
Share of conversions driven through AI on average; usually the second or third channel52:47
aggregateSome app founders the guest spoke to who optimize for AI searchshare50%
approximate
Share of downloads attributed to ChatGPT; depends on the category7:44
benchmarkNumber of options an AI chatbot presents in a recommendationcount3–5 options
approximate
AI answers list 3-5 options; apps not on the list get no visibility8:38
benchmarkHow long content keeps its AI-search visibility without updatestime to milestone3 months
approximate
Content not updated for ~3 months sees a sharp drop-off in AI citations36:07
benchmarkNumber of website content pieces for a baseline of AI-search visibilitycount10–20 pages
approximate
Pieces of content needed for an initial visibility push28:13
Australian e-prescription app (BrandViz client)share5–10%
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)share50%
approximate
Share of AI answers mentioning the app after ~30 days of fixes and content; its website became the most-cited domain41:27
Australian e-prescription app (BrandViz client)count20 content pieces/month
approximate
Pieces of website content published over 30 days41:27
Australian e-prescription app (BrandViz client)count4–5 competitors
approximate
Real competitors in its niche (less contested than most)42:06
Pokemon card grading appcount11K pages
approximate
Website pages, one per Pokemon card and variation26:39
Pokemon card grading appshare50%
approximate
Share of traffic coming through AI (mostly ChatGPT); 'more than 50%'52:47
Atenshare10%
approximate
Share of AI answers to comparison/alternative questions that mention the app by name (BrandViz test)18:41
Atenshare40–50%
approximate
Split of the app's AI citations between App Store listing and its website: first given as 50/50, later as roughly 40/6039:21
aggregateApps cited by AI chatbots for ~1,200 app-blocking category questions (BrandViz test)count136 appsDistinct apps among 315 App Store citations; AI reaches far down the long tail of apps38:46
AppBlockshare70–80%
approximate
Share of AI citations coming from its own website rather than the App Store (also true of other larger competitors)39:21
Freedomcount320 queries
approximate
AI answers mentioning Freedom out of ~600 app-blocker queries run by BrandViz1:03:57
App Masters client fasting apprank1 rankTop app recommended by ChatGPT for fasting7:20
FreedompriceUSD 40In-app purchase price seen on its App Store listing1:15:41
AtenpriceUSD 20Subscription price on its paywall (3-day trial); host thinks it is too cheap1:15:41
AppBlockpriceUSD 30In-app purchase price seen on its App Store listing1:16:05
OpalpriceUSD 90/yearAnnual plan price seen on its App Store listing1:16:05
NutraMindcount6 ratingsApp Store ratings at time of audit1:11:20