How He 3X’d Install-to-Paid in a Niche Fitness App Category

After years failing to scale a general workout app (the host had urged him to niche down), built FaithFull for Christians on a health or weight-loss journey: two niches combined into one micro-niche. The product frames fitness as honoring the body as God's temple rather than pride/ego, adds devotionals, prayer calls and a faith community, and all ads/onboarding copy speak to that identity.

No free trials at all; users must buy to use the app, and the app tries to deliver value from day one. The founder tried repeatedly to make trials work and they never converted as well. Accepts occasional one-star reviews from people who wanted a trial. May test trials again later.

Took the discovery script from a year-long fitness phone-sales mastermind and rebuilt the onboarding in Adapty's no-code onboarding builder (no dev release needed). Structure 'understand, connect, convert': (1) first screen asks a pain question, 'What brings you here today?' with answers like 'I keep starting over', 'Macros confuse me', 'I'm busy and skipping meals'; (2) personal data (height, weight, goal weight); (3) deeper pain: 'What feels hardest right now?' (no time, no motivation, lack of consistency) with a 'that's my struggle' button, 'How long has it been a challenge?' (1-3, 3-5 years), 'How is this affecting you?' (low energy, brain fog, poor sleep, stress, discouragement); (4) future pacing: 'What blessings will you experience when you get there?' and 'What will life look like after you hit your goal?'; (5) show how the app gets them from pain to promise with a you-are-here-to-there graph and 'road to consistent God-honoring habits'.

Launched with a ~10-screen onboarding modeled on the previous app and kept adding screens through tests until it reached 26. The length filters out people who aren't serious and gives room to uncover pain points. January onboarding changes pushed paywall view rate to 62% despite the length.

To justify paying upfront without a trial, asked the app's high-ticket coach (Natalie) to run one monthly group coaching call for every member, included in the membership. Added a single onboarding screen announcing live coaching; nothing else changed (same prices, one variable). Ran as a long Adapty split test until statistically significant.

Starting from $12.99 monthly / $29.99 quarterly / $59.99 annual, split tested monthly at $14.99 with the rest unchanged. Goal is to shift the plan mix to annual, which had the highest LTV in the previous app.

Took the winning paywall and split tested monthly $14.99 -> $19.99 and quarterly $29.99 -> $39.99 with annual held at $59.99, so annual is only ~$20 more than quarterly (about $5/month). Courage came from an accident: after the earlier $12.99 -> $14.99 test he forgot to update the Apple price, which had been $19.99, and iOS monthly conversion hadn't suffered. Now rolled out on both platforms.

After the higher monthly/quarterly prices did well, launched a split test raising annual from $59 to $89 to see whether conversion holds.

The paywall had a 'Smart start' bubble above the quarterly plan. Split tested removing it, leaving the annual plan as the only highlighted option.

10% of all revenue goes to evangelism, and the paywall's third bullet says 10% of your membership goes to spreading the gospel. The annual plan badge reads 'Most impact'. Split tested it against 'Best value' and 'Save X%'. Next test: 'Most popular'.

Early on, tested placing login/account creation later in the flow instead of as the first screen. Moved it back to first because purchases were harder to track and tie to a user otherwise.

Every campaign in the main Meta ad account optimizes for purchases, from day one. Any install or trial campaign test runs in a second ad account, because the account 'remembers' who it was fed and install/trial data can steer future purchase campaigns toward cheap non-buyers. Purchase events carry revenue. Advice came from someone who ran millions in Meta app spend.

Campaigns still optimize for purchase, but the app also sends Meta's standard e-commerce events: fb_mobile_complete_registration on sign-up, fb_mobile_content_view once when the paywall is shown, and fb_mobile_initiated_checkout once when the paywall buy button is tapped. Mapped via Adjust (works with the plain Facebook SDK too). Meta then learns from registration, paywall view and checkout intent before the first purchase arrives; cost per checkout also shows which ads bring quality traffic.

Also fires fb_mobile_achievement_unlocked (e.g. a 7-day food logging streak) and fb_mobile_level_achieved so Meta can learn which buyers stay engaged, not just who purchases.

Constantly launches new creatives because winners fatigue after people see them a few times. Formats: videos, UGC, written testimonials, and a new style where AI turns a written testimonial into a cartoon graphic (10 variations made after the first one got a purchase). Hundreds of ad variations tested over ~8 months.

The top ad was an Arcads AI guy walking in a park with the hook 'If you love God but you're trying to work on your health, you need to see this', then explaining FaithFull. Hired a female UGC creator to copy it exactly: same park setting, same walk, same script and delivery. Generally makes multiple variations of any ad that scales.

Runs creative testing campaigns on Android, where events arrive immediately (cost per install, registration, checkout). Winners go into a single long-lived Android scaling campaign that builds purchase history; new winners are tested against the current champion and switched off if they lose. Only ads that hold up over time are moved into a single iOS SKAN campaign, because SKAN data arrives late and testing there would need ~$1,000/day. Started at $50-100/day, scaled to $150/day once purchases came.

In Adjust: fine conversion values ordered install -> registration -> paywall view -> initiated checkout -> subscription revenue tiers by monthly/quarterly/annual (annual highest; threshold set at $43 USD so $59 CAD annual counts). Coarse values for the first postback (0-2 days): started with paywall view as 'medium' while purchases were scarce; at 100+ installs/day switched medium to any purchase and high to annual. Second (3-7 days) and third (8-35 days) postbacks: medium = monthly/quarterly, high = annual.

Kept Android and iOS campaigns in one Meta ad account. Months of Android purchase data taught the account who the audience is, so iOS SKAN campaigns started with that knowledge.

Split tested the first onboarding screen: a feature message ('Track calories...') vs a pain message ('Struggling to stay consistent even with God'). The same pain copy (struggling to stay consistent, tracking meals without stress, too busy) is used in App Store screenshots and ads.

After the gender question, men and women get entirely separate flows with gender-matched screens. Goal screens use images, not text ('What's your ideal body type?' with four body pictures). Later screens reuse answers: 'What feels hardest right now with toning up?' and the final chart shows their goal weight ('180 lb is just the milestone').

In Oomph, app users were upsold to a sales call and closed on a high-ticket nutrition coaching program. FaithFull now does the same: users upsell out of the app into weekly high-ticket coaching with coach Natalie.

In-app community where members post, get replies, and coach Natalie responds with tips; global coaching calls for all members. In Oomph, community was promised in screenshots but never delivered well in the app.

The last feature added to the Oomph workout app was AI calorie tracking via an SDK. Seeing many users actually use it (fast logging, AI trending), the founder built a new app around it, adding the faith angle.

Ran App Store (Apple Search) ads but found Meta easier to understand and more successful; planning to do more later but time is limited.