How Done grows

For health, fitness and productivity apps, the first week of January is 'the Super Bowl / Black Friday' of results: average cost per download is about half the rest of the year. Spent ~75% of the annual ad budget in the first month; the rest of the year was for testing creators and concepts to rerun in January. January spend set a new MRR/ARR baseline for the company each year.

In 2020, with iOS 14 hurting ad tracking, compared spending money on ads vs buying apps. Bought Done (habit tracker, bundled with 5-6 other apps) at a 'pretty good price' because it was under-monetized, then improved paywalls and in-app subscriptions. Keeping the asset, user base and tech (resellable later) made acquisitions more reliable than ad spend. Repeated for 30-35 apps 2020-2025 across journaling, habits, fasting, money tracking, calendar and mindfulness (Mindfulness App to round out categories). Buying apps was usually the better use of capital, except in January when ad spend won.

Aggregated data across thousands of videos to define what made winners. Briefs (slide decks) split concepts into lifestyle, trend, educational and testimonial; each specified the trending sound, hook type (curiosity, empathetic, reflective), captions, pace, length 10-15s with 25-27 cuts, the number of app shots, and that phone clips are POV looking down at the screen. Angles included morning routine, evening routine, wellness, mindset and vlogs (hook -> habits -> app crossing off habits). Paid someone to scroll TikTok and send a list of trending songs at the start of every week, and updated briefs regularly.

Posted all creator videos organically on brand TikTok accounts (e.g. do.habits, growthbundle). Once a video broke out of the 0-300 view 'jail' to ~5-10K views, or drew 10-100 comments even on few views, boosted it with ad spend, increasing it while it hit CPI/subscription targets. If it really took off, ran it as a Spark Ad on the creator's own account: 4-5 ways to push one video. Organic-first is the cheaper route in quiet periods; in peak season or when time-constrained, launch straight into paid level-1 testing.