How Stronger grows

The app was freemium for a long time. A toggle mistake on New Year's Eve made it a hard paywall for a couple of hours; conversions spiked with few complaints. They then made it trial-first: to use the app you must start a Pro free trial; if you cancel, you get the free version after the trial expires.

The 'fade-in': a ~6-second Instagram Reel that fades in from black to one image (the app's colored strength-score muscle heatmap). Always the same sound (Future's 'Let Me Know'), used like a brand sound. App logo added so viewers know it is a workout tracker. Posted repeatedly for 5-6 months; takes Peter a couple of hours a week to schedule, using internal tooling to produce variants.

Before the app was ready, co-founder Jack filmed a quick video on a work break panning up a Canva-made timeline that converts total workout volume into objects (Easter Island head, Stonehenge slab, '10 Donald Trumps'). It was clearly founder-posted. People followed the page to be told when the app launched; Jack then finished the app in 1-2 weeks.

Commenters posted their lifts ('I benched 220 lb for five reps, what is that?'). Co-founder Jack stitched each comment, typed the lift into the app on camera and held the phone up: 'That's two blue whales / 500 bowling balls'. Viewers kept commenting new lifts, creating an endless loop of content. Guest recommends it again now: a real face on camera stands out against faceless AI content.

Every product decision asks 'how does this look as a screenshot?' and 'is it obvious what it is?'. Bright-colored strength heatmap; weight-lifted milestone badges that total all volume lifted ('you've lifted the Empire State Building / a Boeing 747'), made as cool-looking as possible to feature in TikTok slideshows and carousels. Guest notes the fade-in's weakness: viewers often don't realize it's from a workout tracker, so they added the app logo.

Main TikTok account (run mostly by Peter's brother) posts only 2-3 slide carousels of app screens. Keep the real UI chrome (share-to-Instagram button, download button) so it reads as authentic content rather than an ad; don't over-polish.

Brother runs paywall designs; tests in Superwall. Focus on macro variables first: how often users see a paywall, how hard it is, which features it's attached to and whether it matches the feature's messaging, and format (video paywall vs multi-stage). Only after a macro winner, test micro copy changes. Warns that a 2% difference after a one-week test is noise; micro changes need months of volume.

Inspired by a viral Twitter thread about a web app that turned bench/squat numbers into performance levels, which nobody had built on mobile. Built a strength score: a colored muscle heatmap comparing you to people of the same body weight ('elite chest, novice legs'), so users can claim a status ('I'm an elite lifter') instead of quoting kg. Chosen to stand out in a market investors called a commodity, and to double as the main marketing asset.

Look for replicable formats working in other industries (e.g. a finance app), work out why people engaged, and reskin for your niche. Content buckets: entertain (gym-bro humor) or educate (save-worthy tips). Try a few, watch average watch time, then iterate on hook, angle and audio, treating each video as a funnel (hook -> body -> App Store). Keep a separate phone whose feed is trained only on what the target customer (a ~20-year-old gym bro) would like.

Skipped raising investment; got grant funding from government agencies in Northern Ireland, pitching at startup events.

Once a format wins, scale it horizontally: post it from multiple accounts (own accounts, reposters, burner phones). Make micro-adjustments per post so the algorithm doesn't flag spam: change a couple of words in the hook (AI makes this easy), switch to a light-mode background, change the legend, using Figma templates. A/B test variants to push average watch time up (e.g. from 6 to 9 seconds).

Peter and his brother Joseph make almost all content, with internal tooling and a few affiliates. Agencies tried a little before; incentives don't align and they take ~20% margin. Use them only to learn something you can't, or to quickly stand up volume (e.g. pushing a known winner across 50 accounts) when unit economics are known, then bring it in-house.