The internet has lied to you about podcasts. The lie has three parts. First, that you need an audience before you start. Second, that you need a mic that costs more than your last grocery run. Third, that you need a niche, a brand, a logo, an intro song, and a content calendar before recording episode one. None of this is true. The only thing you actually need is a conversation worth having and a way to record it.
The lie about audience
“Build an audience first” is the most common piece of advice given to new podcasters. It's also the most useless. Audiences come from interesting episodes. Interesting episodes come from doing the thing. You don't practice swimming on land. You don't build an audience by tweeting about a podcast you haven't recorded.
The math is also wrong. The first 50 episodes of nearly every successful podcast have functionally zero listeners. Joe Rogan's first 100 episodes were obscure. Lex Fridman's early conversations were watched by tens of people, not millions. The compounding starts much later than the work does. So start the work.
The lie about gear
You have a microphone in your laptop. You have a microphone in your phone. You have a microphone in your earbuds. Any of them is enough. The point of episode one is not to sound like NPR. The point is to make a thing that exists. A scrappy recording with a good conversation beats a perfectly mastered recording of two people pretending.
After ten episodes, if you're still in love with the work, then buy the mic. After fifty, buy the interface. After a hundred, buy the treated room. By that point you'll know which gear actually solves a problem you have, instead of which gear looks good on a podcast subreddit.
What you actually need
- One guest. A friend, a stranger, a colleague, anyone. You don't have to interview a celebrity. The most underrated podcast format in 2026 is “two curious people talking honestly about a thing they both care about.”
- One question. Not a script. A question. The best podcast hosts don't prep monologues; they prep one good opening question and then listen.
- A way to record. Quicktime, OBS, Riverside, PodRandom — anything that makes a file. Hit record at the start, hit stop at the end. That's the technical requirement.
The 90-minute first episode
Block 90 minutes on a calendar. Not when you feel ready. Now. Find a guest who can talk for an hour about something specific. Open with one good question. Don't edit. Don't add an intro. Don't worry about the title. Just publish it somewhere, even unlisted, even on your own laptop. The point is to ship one. Then you ship a second. Then a third.
If you can't find a guest, that's the actual problem most new podcasters have. It's not gear. It's not branding. It's not audience. It's guest supply. The traditional answer is to network, beg friends-of-friends, send cold DMs. The modern answer is to be matched with someone who likes the same things you do, on a platform built for it. We made one of those at PodRandom.
Where the audience comes from
Audiences arrive when the work is interesting enough to share. Almost every modest podcast hits its first growth inflection between episodes 20 and 50, when one specific episode happens to land in a community that cares. You can't predict which one. You can only guarantee you have enough episodes for one of them to break.
So the only growth strategy that actually works is volume of honest work. Episodes that are 60% as good as your best, but real. Episodes where you ask the question nobody else is asking. Episodes where you publish the conversation that scared you a little. The audience finds those. They don't find your branded intro music.
The compounding part
Around episode 30, something changes. You'll have an archive. The archive is the moat. Each new listener consumes 5–10 back episodes if you're any good. The archive is what makes a podcast a real product instead of a hobby. And the only way to have an archive is to make episodes for the year you don't have one.
This is the part nobody tells you about. Year one is mostly invisible. Year two is your archive recompounding. Year three is when people start saying they've been listening forever.
The hardest unfair advice
Don't niche down before you start. The advice to pick a tight niche is good for marketing departments and bad for new creators. You don't know what you're good at yet. You don't know which conversations excite you yet. Record 20 episodes about whatever you're curious about. Notice which ones leave you feeling like you could record forever. That's your niche. It chose itself.
So what now
Pick a guest. Pick a question. Pick a date this week. Hit record. The first ten episodes will be embarrassing. So is the first month of any new craft. The embarrassment is the cost of admission. Most people aren't willing to pay it. Pay it cheerfully and you'll be doing this for years.
If you're reading this and the guest part is your blocker — the “I don't know who to invite” part — PodRandom matches you with a stranger who shares your interests and you can record the conversation immediately, in your browser, no scheduling or booking. It's not the only way to start a podcast. It's the cheapest one with the lowest excuse-budget. We built it because we hated guest hunting, too.