Most "how to start a podcast" guides are written to sell you a microphone. This one is written on the assumption that your real constraint isn't equipment — it's whether you'll still be doing this in twelve months, when roughly 85% of podcasts have gone quiet.
So we'll cover the gear quickly, because it barely matters, and spend the time on the decisions that actually determine the outcome.
Step 1: Get the Concept Narrow Enough
The single most common mistake is a concept that's too broad. "A podcast about business" is not a concept. "Interviews with people who bootstrapped a business to £1m without funding" is.
Narrow is better for three concrete reasons:
- It makes you findable. People search for specific things.
- It makes sponsorship viable. Niche shows with 10,000–50,000 listeners now deliver better sponsor ROI than mega-shows — but only if the niche is legible.
- It makes episodes easier to produce. A narrow concept tells you what the next episode is. A broad one leaves you staring at a blank page every week.
The test: can you name 20 episodes right now? If not, the concept is too broad, and you'll discover that around episode 6.
Step 2: Choose a Format You Can Sustain When Tired
Be brutally honest here, because this decision kills more shows than anything else.
- Solo — cheapest and most flexible, hardest to make compelling. No scheduling overhead.
- Co-hosted — easiest to make enjoyable, but you've doubled the scheduling problem and taken on a dependency.
- Interview — endless content supply and built-in promotion, but adds booking, prep, and scheduling. See how to book great guests.
- Narrative/produced — highest ceiling, brutally high effort. Not a first show.
The real test isn't whether you can make a great episode when inspired. It's whether you can make an acceptable one at 9pm on a Tuesday when you're flat and busy. Design for that night.
Step 3: Buy Almost Nothing
Here's the whole truth about gear:
- A decent USB microphone (roughly £60–120 / $70–150) is genuinely all you need. Audio quality matters, but the gap between a good USB mic and a professional setup is far smaller than the gap between a bad mic and a good one.
- Wired headphones — any pair, to prevent echo.
- A quiet room with soft furnishings. Room treatment beats equipment. A closet with clothes in it genuinely outperforms an expensive mic in a bare room.
- Free software — Audacity, GarageBand, or your recording platform's built-in editor.
That's it. Do not buy an interface, a mixer, or a second mic until you've published ten episodes. Gear purchases feel like progress and aren't.
If you're recording video — and most top shows are now video-first — add one camera and good lighting. Lighting matters more than the camera.
Step 4: Understand the Real Time Cost
This is where new podcasters get ambushed. Recording is the fun part and the small part.
A realistic hour-long episode:
| Task | Time |
|---|---|
| Planning / prep | 30–60 min |
| Recording | 60 min |
| Editing | 60–120 min |
| Show notes, description, chapters | 30–45 min |
| Clips and social | 30–60 min |
| Uploading and publishing | 15 min |
| Total | ~3.5–6 hours |
That's per episode. Weekly, that's a part-time job.
Two ways to cut it dramatically:
- Edit less. Light editing beats heavy editing for most conversational shows, and listeners care far less about small imperfections than you do.
- Automate the writing. Show notes, descriptions, and chapter markers are the most automatable part of the job. DriftNote's creator tools generate show notes, episode descriptions, and style-matched output from the episode itself — and timestamped summaries effectively pre-scout your clips.
Getting publishing from four hours to ninety minutes is the difference between a show that lasts and one that doesn't.
Step 5: Hosting and Distribution
You need a podcast host (Buzzsprout, Transistor, Captivate, RSS.com, Spotify for Creators, and others). They store your audio and generate the RSS feed that Apple, Spotify, and everyone else pull from.
- Expect £10–25 / $12–30 per month. Free tiers exist with storage limits.
- Submit your RSS feed to Apple Podcasts, Spotify, and YouTube. Everything else largely follows automatically.
- Publish to YouTube properly — real title, thumbnail, chapters, description. It's the biggest remaining discovery channel, especially as AI search erodes traditional referral traffic.
Don't agonise over the host. They're broadly similar and switching is straightforward.
Step 6: Launch With a Buffer
Record three to five episodes before you publish anything. This is the most-ignored advice in podcasting and the most important.
A buffer does three things: it lets you launch with several episodes so new listeners can binge, it means your first published episode isn't your first-ever recording, and it protects you from the illness or work crunch that otherwise breaks your schedule permanently.
Launch with three episodes at once, then settle into your cadence.
Step 7: Set Expectations That Won't Break You
Most shows grow slowly and then compound. Ninety downloads on episode ten is normal, not failure.
Measure things that predict survival:
- Episodes published (the only metric that correlates strongly with eventual success)
- Whether you'd proudly send a stranger episode 12
- Listener replies and messages
Ignore, for the first six months: download rankings, comparisons to established shows, and the chart.
And answer this honestly before you start: would you still make this if it had 200 listeners in a year? Most shows that survive are made by people whose answer is yes.
The 90-Minute Version
If you want the whole guide compressed:
- Pick a concept narrow enough to name 20 episodes.
- Buy one USB mic. Nothing else.
- Record in a soft, quiet room.
- Record three episodes before launching.
- Automate show notes and descriptions.
- Publish to Apple, Spotify, and YouTube.
- Commit to ten episodes before judging anything.
Where to Go From Here
- Try DriftNote's creator tools
- 85% of podcasts are dead — here's why
- How to write podcast show notes with AI
- Podcast monetization in 2026
Starting is easy and everyone does it. Reaching episode 50 is rare and is almost entirely a function of how sustainable you made the process on day one. Build for the tired Tuesday, and the rest takes care of itself.
Prices are approximate 2026 ranges and vary by region and retailer.