MAU Vegas 2026 Debrief: AI, Growth & What Actually Worked

Bootstrapped founder spends roughly $300/day on Meta web-to-app campaigns and ~$100/day on Meta app install campaigns (about 75% of Meta spend to web-to-app).

The previous owners followed Steve's ASO playbook: built multiple apps, and one took off. It ranks #1 for a few keywords with popularity ~41 plus several smaller keywords that add up, with 'a hint of' Apple Ads, and mostly sells in the US. That yields ~1,000 downloads a week (4-5K/month).

App has no keyword volume, so Meta (App Referrer source) is the primary acquisition channel. Once Meta spend started, App Store search downloads also rose because more people searched for the brand.

Create a Max Conversions campaign. Set the target cost per acquisition (e.g. $2) and a daily budget of 15x the target (~$30/day) instead of Apple's recommended 30x. Add an ad group with the four highest-traffic generic keywords; if there are none (e.g. dating), use the 4-5 biggest competitor brands (Tinder, Bumble, Hinge). Audience: 'Choose specific audiences' -> 'New customers' (only shows to people without the app; does not cut impressions). No age, gender or location targeting. Compared against App Masters' internal 'Basic Plus' campaign.

Do not add gender, age or state-level targeting to Apple Ads campaigns; Apple's privacy rules make them sharply limit impressions. Targeting 'New customers' only is fine.

Whole Meta ad account optimized for purchases (the app has no trial; start trial works too). Also send Meta the pre-funnel standard events: registration (if any), ViewContent fired on paywall view, and InitiateCheckout, so Meta gets more intent signal. Borrowed from e-commerce playbooks.

Client with ~1M monthly active users had no paywall in onboarding; App Masters added one. Steve's rationale: users who subscribe are more likely to retain.

Fitness app, first email after a trial starts: subject line changed from 'The benefits of HIIT working out' to 'Our most popular workouts'. Same idea in-app: a 'Popular' section (like Calm's 'Popular on Calm') to guide users through a large content library.

In the onboarding 'How did you hear about us?' question, add 'TV ad' (put it at the top) and an authority referral option such as 'lawyer' or 'referred by financial adviser', even if you run no TV ads. Cal AI, Resumee and Flow show similar TV/YouTube options.

Old flow: onboarding screens -> forced sign-up -> more onboarding screens -> paywall. New flow: swap them so the paywall comes before registration; ask for the account after purchase ('to unlock premium we need your info'). Watch the install/first open -> paywall displayed rate, especially for simple utility apps; long onboarding suits health/fitness/education.

Steve's friend debated a hard paywall vs a soft paywall with more content locked, and tested both; he also locked up more content.

When a user says no to the subscription paywall, instead of dropping them on the home screen, show 'Unlock your first three on us for just two bucks' as a one-time in-app purchase (about $2-3). Tripwire: a small first purchase makes a later subscription more likely.

Blinkist's paywall change, published as 'How solving our biggest customer complaint led to a 23% increase in conversion'; the model for trial-reminder paywalls later copied by Resumee and Cal AI. No further detail given in the episode.

Two yearly options on the paywall: yearly at $59.99 with a 7-day free trial, and a paid intro offer of $1.99 for the first 30 days that renews at the $59.99 yearly price, replacing the monthly plan. Steve says it works well for most clients; Resumee uses it too.

First tried a Resumee-style screen before the paywall ('we want to offer 7 days free, you'll get a reminder'); it underperformed. Then switched to two screens before the normal paywall: 'We want you to try [app] for free' with benefits as a still image or video, then 'Try for $0, no payment due' with social proof, then the usual paywall, changing nothing else.

Default trial lengths: 7-day trial on yearly plans, 3-day on weekly plans. When a user closes the normal paywall, show a follow-up offer with a longer trial. Steve contrasts this with RevenueCat's macro data that longer trials convert better overall.

Check your app type's keyword popularity in any ASO tool (MobileAction, AppFollow). Popularity 30+: use ASO and Apple Ads (pull demand). No search (e.g. social apps): use Meta to push the message. Then focus on one or two channels and exhaust them before adding more.

Before building, check the keyword: if the top apps all have it in their title and huge rating counts (e.g. 'meditation': Calm, Headspace, Insight Timer with ~1M ratings), skip it. Pick keywords where #1 is reachable quickly, build the app, rank #1 to find early product-market fit, then scale with Meta.

Create content answering the questions people ask AI assistants about the app: 'What is Momingo? Does Momingo work? Can I use Momingo? Does Momingo support Washington, D.C.?' Steve suggests the same for feature questions ('Does BrainJams support language learning with spaced repetition?').