How to launch and grow a podcast from scratch

Executive overview

Most podcasts die early — only 150,000 of 4 million have published 10+ episodes in the last 10 days. Ten consistent weekly episodes puts you in the top 4% of all podcasts ever created.

Pick a topic you'd discuss for free indefinitely. Without that, the inevitable revenue-free months become unsustainable.

The single biggest lever is content quality — treat it like product-market fit.

Deciding whether to start

• Only 150k of 4M podcasts are active (10+ episodes, published recently) — consistency alone is a competitive advantage
• Start with a committed "season one" of 8 episodes; no obligation to continue if it doesn't feel right
• Expect 6–9 months with no revenue and far more work per episode than anticipated (prep, editing, guest outreach, show notes)
• Choose a tight niche — the harder it is to explain, the harder it is to grow
• Ask: what topic makes people lean in at a dinner table and text you afterwards?

Launching for traction

• Launch with 2–3 episodes ready, not one — prevents scrambling and signals commitment
• Apple charts rank on momentum of new subscribers, not total downloads — a spike at launch earns a screenshot you can use forever ("top 10 podcast")
• Use that screenshot immediately: pitch dream guests with "check the charts right now"
• Build any existing audience (newsletter, friends list) before launch to seed the spike
• Record 5 episodes before publishing — pick the strongest as episode one

Growing an audience without a platform

• Build a "friends newsletter" now, even before you launch — even 1,100 casual subscribers from life are an asset
• Find communities (Reddit, Twitter, forums) where your niche is discussed; answer questions as an expert, let your bio do the rest
• Create clips for platforms with built-in discovery (TikTok, YouTube Shorts, Instagram Reels) — drives brand awareness even if direct download attribution is unclear
• Record in-person video for clips where possible — guests are far more likely to share high-quality clips to their audiences
• Go on other podcasts as a guest; podcast listeners are the best target demo for a new podcast
• Craft guest pitches that prove you understand their show — generic outreach fails

Content and consistency

• Be someone's favorite podcast, not everyone's acceptable one — Tim Ferriss's most niche episodes rank among his best
• Survey listeners around episode 50: almost every episode will be someone's favorite
• Release consistently and at a fixed time — listeners build rituals around it
• Experiment with format: solo episodes, Q&A, narrative, co-host — find what feels natural
• Repeat the show's premise at the top of every episode so listeners can describe it to others in one sentence

Podcasting stack

• Mic (entry): ATR 2100X (under $100, USB + XLR)
• Mic (upgrade): Shure SM7B or MV7
• Recording: Riverside
• Editing: Descript (transcript-based editing; works for amateurs and pros)
• Audio interface: Focusrite Scarlett 2i2 (for XLR mics)
• Hosting: Simplecast (affordable entry tier, scales to monetisation)
• Website: Podpage (auto-generates from RSS feed, updates automatically)
• Analytics: Chartable (tracks cross-promo attribution); PodStatus (chart rankings)

Benchmarks and metrics

• 3,000 downloads per episode ≈ top 1%
• 10,000 downloads per episode = taken seriously by networks and sponsors
• Average listener completion rate is below 50%; 65–70% is excellent
• Focus on directional growth (episodes 4–6), not launch numbers inflated by social capital
• Cost to acquire a podcast listener via paid ads: ~$3–$10; only viable once ad revenue justifies it
• Overcast ads: cheap, measurable (click-through + subscription rate) — useful as a low-cost content/positioning test