The question is almost always triggered by one number. There are 4.7 million podcasts in the public index, and that figure makes a new show sound pointless before it exists. The number is real. It is also the wrong one to be frightened of.
On 16 September 2026 the Podcast Index held 4,725,462 podcasts. In the previous 30 days, 339,452 of them published an episode. That is 7.2 percent. The market a new show actually competes in is roughly one fourteenth of the number that makes people hesitate, and it is a market where the average competitor has already stopped.
So no, it is not too late, and that answer is too simple to be useful on its own. What follows is the evidence, including the parts that argue the other way, and then the harder question underneath: too late for what.
The podcast index holds 4.7 million shows and 339,452 of them published last month
The saturation argument treats every listed show as a competitor. Directories do not delete dead feeds, so the total count measures history rather than activity. Reading the publishing windows instead of the total changes the picture completely.
| Publishing window | Shows that published | Share of the 4,725,462 in the index |
|---|---|---|
| Last 3 days | 94,615 | 2.0 percent |
| Last 10 days | 228,941 | 4.8 percent |
| Last 30 days | 339,452 | 7.2 percent |
| Last 60 days | 407,771 | 8.6 percent |
| Last 90 days | 471,507 | 10.0 percent |
Nine in ten listed podcasts have not released an episode in 90 days
Subtract the 90 day figure from the total and 4,253,955 shows in the index have published nothing in three months. That is 90 percent of the catalogue sitting still. A weekly show that publishes for a year is not joining 4.7 million competitors. It is joining roughly 340,000, and it is doing something that most of that 340,000 will stop doing.
Four in ten podcasts launched in 2026 had already stopped by the middle of the year
Podchaser pulled its database on 30 June 2026 and looked at every show with a first episode dated between 1 January and 30 June 2026. Of the 153,767 podcasts that launched in that window, 64,100 were already inactive when the data was pulled. That is 41.7 percent, gone inside six months, most within weeks of the debut.
Two readings sit inside that number and both are worth holding. The discouraging one is that podcasting is harder to sustain than it looks. The encouraging one is that persistence is unusually valuable here, because the field thins out on its own. Publishing episode twenty is a competitive advantage in a category where most entrants never reach episode six.
New podcast launches fell from their 2020 peak for a reason that is often misread
This is the statistic people quote when they argue that podcasting is in decline, so it deserves the full version rather than the headline.
Listen Notes data reported in industry coverage put 2024 launches at 184,603, down 21 percent on the 232,670 recorded in 2023, and far below the pandemic peak of roughly 920,000 in 2020. On its own that reads as collapse. The same dataset showed active shows moving the other way, from 260,690 in 2023 to 598,081 in 2024.
The reason both numbers moved at once is structural rather than cultural. Podnews has documented the mechanics: a major free hosting platform stopped automatically submitting new shows to Apple Podcasts in mid 2021, which removed a large volume of single episode feeds that had inflated earlier counts. Submissions from free hosting fell by around 93 percent in the comparable period, while paid platforms fell only 10 to 20 percent, and the number of new episodes published each year stayed comparatively flat.
What disappeared was abandoned experiments, not shows with listeners. The launch count fell because the counting changed and the noise left. Anyone starting now is entering a smaller, more serious field than the raw 2020 figure suggests, which is a better position than the one the headline describes.
Australian podcast advertising reached a record quarter in June 2026
Audience interest is one signal and commercial investment is a firmer one. The IAB Australia Internet Advertising Revenue Report for FY26, compiled by PwC Australia and released on 31 August 2026, put the Australian internet advertising market at $19.8 billion, up 14.0 percent year on year. Inside that market, audio advertising grew 6.6 percent to $353 million, and the growth was driven by podcasting.
| Measure, Australia | FY26 result | Change |
|---|---|---|
| Internet advertising market | $19.8 billion | Up 14.0 percent year on year |
| Audio advertising | $353 million | Up 6.6 percent |
| Podcast advertising within audio | 40 percent of total audio spend | Up 10.1 percent |
| Streaming audio within audio | Remainder of audio spend | Up 4.3 percent |
| Podcast advertising, June 2026 quarter | $40.0 million, a new quarterly peak | Up 13.1 percent on June 2025, up 22.6 percent on the March quarter |
A 40 percent share of a $353 million audio market puts Australian podcast advertising at roughly $141 million for the financial year, and the June quarter was the strongest on record. Advertising money is slow and conservative. It does not set new peaks in a format it has decided is finished.
Australian podcast listening grew again in 2026 and video accounts for the increase
The Infinite Dial Australia 2026 study reports that 55 percent of Australians aged 10 and over consume podcasts monthly, up from 52 percent a year earlier. The study surveyed 1,600 Australians aged 10 and over, with fieldwork between 12 and 28 April 2026. Trade coverage of the same study reported that audio only monthly listening crossed 50 percent for the first time, which places roughly 12 million Australians in the monthly audience.
The gap between those two figures is the whole story of 2026. Audio only listening sits at 50 percent. Listening including video sits at 55 percent. Five percentage points of the Australian podcast audience is watching rather than only listening, and that is the segment that did not exist when the market last looked crowded.
Video is the structural change that separates a 2026 launch from a 2020 one
The global figures make the same point at scale. YouTube announced one billion monthly active podcast users in 2025, and reported that viewers watched more than 700 million hours of podcasts on living room devices in October 2025, up from 400 million hours a year earlier. That is a 75 percent increase in podcast watching on television screens in twelve months.
A show that publishes audio only is not competing for that attention at all. This is the single most useful fact for anyone weighing up a late entry, because it means the crowded part of the market and the growing part of the market are not the same part.
| What decided a show in 2020 | What decides a show in 2026 |
|---|---|
| Being early to a topic | Being watchable as well as listenable |
| Audio quality good enough to finish a listen | Framing, lighting and a cut that holds attention on a screen |
| Discovery through podcast apps and charts | Discovery through video platforms, clips and search |
| One audio file per episode | An episode plus short form cuts from the same recording |
| Volume of shows as the barrier | Production standard as the barrier |
The barrier moved rather than rose. In 2020 the hard part was being noticed among a flood of new shows. In 2026 the flood has receded and the hard part is looking and sounding like something worth watching, which is a problem with a known solution rather than a matter of timing.
Too late for what? The answer changes with the goal
The honest answer is not uniform across every ambition, and a guide that says yes to all of them is not being straight with you.
| The goal | Is it too late | What actually decides it |
|---|---|---|
| A general interest show competing for the national charts | Effectively yes. Chart positions are held by broadcasters, networks and established names with promotional reach a new independent cannot match | Existing audience, not start date. Arriving in 2019 without one would not have changed this |
| A niche show for a defined audience | No. Niche categories are where the active 340,000 thins out fastest, and specific beats broad in a market this size | Whether the niche is narrow enough that you can name the listener |
| A business, brand or founder show | No, and the question barely applies. Success is measured in conversations, credibility and sales cycles rather than download rank | Whether the show is tied to a commercial outcome and produced to a standard the brand can stand behind |
| A video first show | No. This is the newest part of the market and the least crowded, with living room watching up 75 percent in a year | Whether you can produce video of a standard that survives a full screen |
| A show you will publish twice and abandon | The wrong question. 41.7 percent of 2026 launches were inactive inside six months | Whether you have the format and the time before you record episode one |
Six conditions decide whether a 2026 launch works
Start date is not on this list. These are, in the order that they tend to break.
- A defined listener. Not a topic, a person. The narrower the description, the easier every later decision becomes.
- A format that survives episode twelve. Shows built on a novelty or a single guest type run out of material before they find an audience.
- Video from the first episode. Retrofitting video later means a back catalogue that cannot be clipped and a channel that starts from nothing.
- A production standard that does not distract. Room reverb, uneven levels and flat lighting are read as a signal about the content before a word lands.
- A publishing rhythm you can hold for a year. Consistency is the variable the attrition data keeps pointing at.
- A reason to continue that is not download numbers. Early numbers are always small. A show tied to a business outcome survives that period. A show tied to vanity metrics usually does not.
Four of those six are decisions. Two of them are production problems with a price attached, and what podcast studio hire actually costs in Sydney prices them honestly, including the point at which a cheaper room produces a more expensive episode.
Konnect Studios removes the two conditions that stop most new shows
Of the six conditions above, the two that most often end a show are production standard and publishing rhythm, and both are what a built studio exists to solve. Konnect Studios is a podcast recording studio in the Parramatta CBD, six minutes on foot from Parramatta station and the light rail, and teams book a Parramatta podcast studio for a first season from $174 an hour on the Capture session.
| The condition | How a session at Konnect Studios addresses it |
|---|---|
| Video from the first episode | Two Sony FX30 cameras record 4K on every session, framed and colour matched before recording starts. Video is the default rather than an upgrade, so the back catalogue is clippable from episode one |
| A production standard that does not distract | A treated recording room at minus 40 dB isolation, up to four Shure SM7dB microphones each on their own channel, and a full lighting rig. A studio operator sets the room before you arrive and runs the session |
| A publishing rhythm you can hold | Sessions book in 1, 2, 4 and 8 hour blocks, so a full day batches several episodes in one sitting. Batching is the practical answer to the consistency problem the attrition data describes |
| A format that survives episode twelve | Two sets, The Vault and The Luxe, inside a single booking. A second look without a second location gives a series somewhere to go as the format develops |
| The editing hours | Blackmagic ATEM live switching cuts angles during the session, so the edit starts from a director's feed. Craft and Conquer sessions add the full episode edit, five reels and one teaser |
Hiring is not the only route and this guide is not pretending otherwise. Free podcast studios in Sydney and who qualifies covers the council and university rooms that suit a first season on no budget, and how to start a podcast in Australia costs the buy your own equipment path in Australian dollars alongside the requirements that apply here.
The numbers point to three decisions rather than a yes or a no
Read together, the evidence says the timing question is settled and the execution question is not. Three things follow from it.
- Narrow the show before you buy anything. The active field is 340,000, not 4.7 million, and a specific audience is how you find your share of it.
- Record video from the first episode. Five percentage points of the Australian audience sit on the video side, and YouTube counted more than 700 million hours of living room podcast watching in October 2025 alone. A back catalogue recorded without cameras cannot be retrofitted.
- Plan for twenty episodes, not for one. The attrition data is the clearest finding in this article, and batching several episodes in a single session is the most reliable defence against it.
Questions people ask before starting a podcast in 2026
Is it too late to start a podcast in 2026?
No, and the number that makes people ask is misleading. The Podcast Index listed 4,725,462 podcasts on 16 September 2026, and 339,452 of them had published an episode in the previous 30 days. That is 7.2 percent. Ninety percent of listed shows have released nothing in three months. The field a new show competes in is about 340,000, and roughly four in ten shows launched in 2026 had already stopped publishing by the middle of the year.
Is the podcast market saturated?
The directory is saturated. The active market is not. Directories never delete dead feeds, so the total count records every show ever launched rather than every show still running. Measured by publishing activity instead, the market is about one fourteenth the size of the headline figure, and it thins further inside any specific niche.
Are fewer people starting podcasts now?
Launch counts fell from a pandemic peak of roughly 920,000 in 2020 to 184,603 in 2024 on Listen Notes data, but the reason is largely mechanical. A major free hosting platform stopped automatically submitting new shows to Apple Podcasts in mid 2021, which removed a flood of single episode feeds. Free hosting submissions fell about 93 percent while paid platforms fell only 10 to 20 percent, and annual new episode numbers stayed comparatively flat. Active shows rose from 260,690 in 2023 to 598,081 in 2024 on the same dataset.
Is podcasting still growing in Australia?
Yes on both audience and revenue. The Infinite Dial Australia 2026 study puts monthly podcast consumption at 55 percent of Australians aged 10 and over, up from 52 percent a year earlier. IAB Australia reported audio advertising at $353 million for FY26, with podcasting up 10.1 percent and now 40 percent of audio spend, and podcast advertising hitting a record quarterly peak of $40.0 million in the June 2026 quarter.
Do you need video to start a podcast in 2026?
Not to publish, but it decides how discoverable the show is. YouTube reported one billion monthly podcast users and more than 700 million hours of podcast watching on living room devices in October 2025, up from 400 million a year earlier. In Australia, monthly listening is 50 percent for audio only and 55 percent including video. An audio only show forfeits that difference, and a back catalogue recorded without cameras cannot be clipped later.
How many episodes should you plan before launching?
Plan a season rather than an episode. Of 153,767 podcasts launched in the first half of 2026, 41.7 percent were inactive by 30 June. Recording several episodes in one session before the first release is the most practical protection, because it separates publishing from the pressure to record every week.
Is it too late to start a business podcast?
Business shows are the case where the question barely applies, because the measure is not download rank. A show that reaches the right several hundred people in an industry can carry a sales cycle, and Australian advertising data shows commercial confidence in the format at a record high. The deciding factor is whether the production standard matches the brand rather than whether the category is crowded.