How Noah Kagan spent $222,000 growing a YouTube channel in 2020

Executive overview

Growing a YouTube channel through marketing experiments is expensive and mostly inefficient. Noah Kagan spent $222,038 in 2020 and found that 90% of new subscribers still came from YouTube's own recommendation algorithm — not paid tactics.

The lesson: marketing amplifies good content but can't replace it. Run experiments, cut what doesn't work, double down on what does.

Great videos are the only sustainable growth lever — everything else is amplification.

YouTube ads

• Best approach: run ads to existing top-performing subscriber-generating videos, not promotional ads
• Make the video unlisted before running ads so original stats aren't affected
• Cost: ~$2.40 per subscriber over $15,000 spent
• Only run ads if you have a clear ROI path — measure cost-per-subscriber against lifetime customer value
• Don't advertise without a way to recoup the spend

Giveaways

• Tesla giveaway (off-platform): $60,000 spent for ~3,000 subscribers ($20/subscriber); half unsubscribed the next day
• Off-platform giveaways attract freebie-seekers, not genuine fans
• On-platform giveaways perform far better:
  • iPad given away during a live collab on Oli Abdal's channel: 800+ subscribers at $0.41 each
  • MacBook Pro given to own audience at 100k milestone: 400 subscribers at $3 each
• Key rule: do giveaways within YouTube if the goal is YouTube subscribers
• If you can't afford prizes, partner with brands who can fund them

Collaboration videos

• Traditional collabs (co-creating videos for each other's channels) generated ~200 subscribers at $0 direct cost but significant team time
• Appearing on other channels is more effective when you're starting out
• The pitch must be value-first: what does the other creator get from it?
• Guest appearances on others' live streams with a giveaway incentive outperformed standalone collab videos

Sponsoring other YouTubers

• Sponsored Furious Pete to eat tacos for a testicular cancer charity fundraiser: $5,800 total, ~300 subscribers (~$20/subscriber)
• High cost per subscriber, but built a genuine relationship and supported a cause
• Creative sponsorships that align with the creator's passion generate better content than transactional deals

Email swaps

• Cross-promote each other's content to each other's email lists at zero cost
• Generated 100–400 subscribers per swap
• Works best when staying in the same medium — email promoting a newsletter beats email promoting a YouTube channel
• Partners included The Hustle, Morning Brew, and similar newsletters

Making great videos consistently

• 90% of new subscribers came from YouTube recommending videos organically
• Three videos per week for seven months = 84 videos
• Top subscriber-generating video (MrBeast earnings breakdown): 5,600 subscribers
• Emailing the video list every Monday and Wednesday to existing email subscribers caused visible subscriber spikes on upload days
• Team cost for this output: $140,000+ in salaries in 2020 alone (~$2,000 per video)
• The channel made ~$5,000/month from YouTube — the operation ran at a loss as a long-term investment

Key takeaways

• Always be testing (ABT): most experiments fail; success comes from running many bets and scaling what works
• Great content is non-negotiable — marketing gets people in the door, quality keeps them
• Commit to the "law of 100": publish 100 videos before judging results; improve incrementally each week
• You don't need $200,000 to start — begin with a phone and reinvest earnings as they come