Ever wondered how the (podcast) sausage is made?
In this slightly "meta" episode, I reveal the step-by-step process to make an episode of English Learning for Curious Minds.
[00:00:04] Hello, hello, hello, and welcome to English Learning for Curious Minds, by Leonardo English, the show where you can listen to fascinating stories and learn weird and wonderful things about the world at the same time as improving your English.
[00:00:21] I'm Alastair Budge, and today we have an unusual type of episode.
[00:00:27] I’m not going to talk about a person’s life, a historical event, or some other weird and wonderful thing going on in the world.
[00:00:35] Today, I’m going to talk to you about how I choose, research, write and produce podcast episodes, and how this process has changed over the past almost six years of doing so.
[00:00:49] I’ve had quite a few requests for this type of episode, and a few weeks ago, I got it again from an interesting Brazilian member called Itamar.
[00:00:58] So, Itamar, and to everyone else listening to this, I hope you’ll enjoy this slightly meta type of episode. I’m also going to share a bunch of things I found out about the kind of people who listen to this podcast from a recent survey, so I think you might find that interesting, too.
[00:01:16] OK then, let’s get right into it.
[00:01:20] Depending on the way you calculate it, there are almost 5 million different podcasts.
[00:01:27] Not individual podcast episodes; 5 million podcasts, 5 million different shows.
[00:01:35] As I’m sure you know, there are podcasts about everything: comedy, TV shows, business, football, knitting, fishing, obscure historical periods, and, lest I forget, podcasts that help you learn a language.
[00:01:52] English Learning for Curious Minds–this podcast–is one of those.
[00:01:57] Of course, it isn’t the only one. There are tonnes of different podcasts that help you learn English, and they tend to follow similar formats.
[00:02:08] Some are conversation-based, with two teachers having a conversation about anything from their personal lives to a particular element of English grammar or language.
[00:02:20] Others might be one person giving motivational advice, or giving what’s essentially an English lesson in an audio format.
[00:02:30] This one is a little different.
[00:02:34] My approach right from the start has always been to present an interesting and thought-provoking subject in a way that allows you to learn English through listening to the episode, and allows you to learn weird and wonderful things about the world at the same time as you improve your English.
[00:02:56] This is now episode 565, so I have had quite a bit of practice making them.
[00:03:03] And in this meta episode, I’ll tell you what I’ve learned and what the process of making an episode looks like now.
[00:03:13] So, let’s go back to the start.
[00:03:16] When I first started, almost six years ago now, I had absolutely no idea what I was doing.
[00:03:24] I had quite a good idea of what I wanted the finished product to look like–a 15- to 20-minute episode that was both interesting and helped people improve their English–but I didn’t have much of an idea about how to actually make one.
[00:03:41] So…I just started, and figured I’d make it up as I went along.
[00:03:48] I chose some topics that I thought might be interesting, I did a bunch of research into them, and I pulled up Google Docs and just started typing.
[00:03:59] When I thought they were good enough, I recorded them using a microphone I borrowed from a friend.
[00:04:06] Then I listened to them again and…I thought they sounded awful.
[00:04:12] Firstly, like most people, I really didn’t like the sound of my own voice.
[00:04:18] And secondly, I thought the content of the episodes themselves sounded dull. I wouldn’t want to listen to them, so why should anyone else?
[00:04:29] In fact, I never released these early episodes. They were on topics I now realise were pretty difficult to make interesting in an audio-only format: there was one on how black holes work, and another one on how the international bond market works.
[00:04:48] Interesting and important subjects, sure, but quite hard to turn into an interesting topic for someone at B1 or B2 level.
[00:04:58] So, I scrapped them - they simply never saw the light of day.
[00:05:04] I decided to choose some slightly easier subjects to tackle, and things that I was more familiar with, and got cracking.
[00:05:14] That was in December 2019, coming on 6 years ago now, and I have just kept going, with a few changes and improvements along the way.
[00:05:25] So, how does it work now?
[00:05:29] Well, first comes the idea for the topic.
[00:05:33] If you are a seasoned listener to this show, you’ll know that we’ve covered a lot of different subjects.
[00:05:40] There’s been everything from episodes about the life of disappeared English aristocrats through to stories about the Vikings, how recycling works, conspiracies from ancient times, the list goes on.
[00:05:54] There are only really three things that need to be true for me to decide that they are worthy of an episode.
[00:06:02] Firstly, I need to find the subject interesting. Whether that means it’s an interesting story about a period of history, about someone who lived a particularly interesting or important life, or about a particularly interesting concept, it needs to be interesting to me.
[00:06:23] Secondly, it shouldn’t be about current affairs or a topic that will become irrelevant in weeks or months to come. My hope is that anyone can go through our library of episodes and choose one that we might have made several years ago, but it will be just as relevant today as it was when it was made. So something like King Henry VIII or Cleopatra - realistically, that isn’t going to change much, so it makes for a good candidate.
[00:06:57] And thirdly, it shouldn’t be about a highly controversial or divisive topic. I try to be as apolitical as possible in this show, within reason, and there are some topics that I know with complete certainty would end up flooding my inbox with angry messages on both sides of the argument. I’m sure you can imagine what might fit into that category.
[00:07:23] There are a few other things I take into consideration too, like whether a story can be properly told in an audio-only format. Some, like if you are talking about works of art or geographical borders, just don’t work quite so well without images, so I’ll also take that into consideration, and that’s the reason there aren’t many episodes about art.
[00:07:49] We also now make one three-part mini-series every month.
[00:07:54] Sometimes these will be about three different subjects that are all on a similar theme, but sometimes there will be a story that I just don’t think can be properly told in a single 20-minute episode, like the one on the Cambridge Five spy ring. This started as a standalone, single episode, but I ended up splitting it into three different episodes, a three-part mini-series.
[00:08:23] So, once a topic is chosen, what next?
[00:08:27] Well, I’ll usually try to get as much of an understanding of the story or the topic as possible before writing anything down.
[00:08:37] This might start with something as basic as skimming over a Wikipedia entry to try to get an overview of the topic, before going into more detail about particular events or parts of the story, reading opinion pieces, downloading datasets and trying to make sense of them, reading books on the topic, it can be anything that helps me understand the topic better.
[00:09:01] Sometimes I start from a position of relative knowledge, if I’m writing about something I already know about. I started this show while living in Malta. I lived there for seven years and made several episodes about things related to Malta, so I was able to write most of those without too much prior research.
[00:09:24] And I’ll often make episodes about people or historical events I’m already reading about, so I’ll be reading a book for fun, taking notes, and then when it comes to writing the episode of English Learning for Curious Minds, I just refer back to my notes, and the process is a whole lot simpler.
[00:09:45] Once I feel relatively comfortable with my own knowledge of the subject, I’ll make a rough plan and start writing.
[00:09:54] A rough plan for me is a very simple collection of five or so bullet points, with general ideas of how I want the episode to develop. Sometimes it’s easy, like if it’s about the story of someone’s life, which typically follows a chronological order, but sometimes it’s harder, and it takes a bit more time and thought to get it right.
[00:10:20] Then, I start writing.
[00:10:23] When I first started, I would write out general bullet points rather than the entire script, but when it came to recording, I found this meant I spent more time pausing and repeating certain parts in a different way, so now I write out the entire script, and I simply read it out loud.
[00:10:46] But I write as I speak, or at least I try to, so my hope is for an episode to sound more like a conversation, more like natural speech, than a formal written article read aloud.
[00:11:02] As you can probably tell, I use a range of different vocabulary and expressions. There is no particular grand science to this, but I try to use language that someone at a B1 level could understand at a bit of a stretch, language that someone at a C1 level could understand almost entirely, but that anyone at B1 to C1 level will learn from.
[00:11:31] This means not artificially dumbing down my language or using simple terms, but I might say something twice in two different ways — first naturally, and then in another, simpler way that will help you understand if you didn’t the first time.
[00:11:52] I speak more clearly than I would if I were speaking to a native speaker, and of course, I make sure that all of the more complicated words and phrases are clearly defined in our transcripts and study packs.
[00:12:06] There are around 100 of these advanced words and phrases per episode, so that’s more than 50,000 different definitions created, and when you remove the duplicates, the ones that have appeared twice or more, there’s something like 20,000 B2 and above words that you’d learn if you listened to every episode.
[00:12:29] Now, in terms of length, I aim to make each episode around 20 minutes long, with some coming in slightly under that and others a little over, which means the script should be between 2,500 and 3,000 words long.
[00:12:49] I try to write the script in one go, and this writing process typically takes anywhere from two to five hours. It’s rarely less than two, sometimes it can be much more than 5, but it tends to fall into that two-to-five-hour bracket.
[00:13:09] Now, you might think this is quick, or you might think it’s a long time, but all I’ll say is that it used to take me a lot longer. After doing it 565 times and writing well over a million words, I've had quite a lot of practice, so it would be strange if I hadn’t got quicker.
[00:13:31] And this brings me on to another important point.
[00:13:35] At various points over the years, I have tried a bunch of things to speed up the process.
[00:13:42] I've experimented with hiring external scriptwriters to write episodes of English Learning for Curious Minds. I posted a job for this once and got more than 500 applicants, and I trialled 30 different ones.
[00:13:58] Unfortunately–or perhaps fortunately–I never managed to find someone who was able to write to a level of quality that I was happy with, and that I felt was true to my voice and style, the voice and style of this podcast.
[00:14:14] In every case, when I would review the script that they had written, I found myself making so many edits and changes to their drafts, and starting to record an episode and realising that something they had written didn’t sound like me, that I realised it would have been quicker for me to just write everything from scratch.
[00:14:37] So that is now what I do: no external scriptwriters, everything on this show is written by me, Alastair.
[00:14:46] I do, however, have another scriptwriting colleague, in the form of various AI models I use to help me at different points in the planning and writing process.
[00:14:58] Specifically, I primarily use ChatGPT, Claude and Grok, but I’ll sometimes use other tools as well for a second or third opinion on something.
[00:15:09] During the planning stage, I will often tell ChatGPT what my proposed outline is, and ask it to point out areas I hadn’t thought about.
[00:15:20] If I get writer’s block, and I’m just not sure where to go next in a script, I might paste in what I’ve written thus far and ask it to give me three different routes to take it next.
[00:15:34] I don’t always use what the AI suggests, and if I do use them, I often find myself having to rewrite much of what ChatGPT provided, but they can be useful for giving me a fresh idea and helping me get back on track.
[00:15:52] I have, by the way, tested multiple times trying to get ChatGPT to actually write episodes in my voice. It does a better job than most of the human scriptwriters I've previously worked with, but it still sounds far too artificial, too unlike me, so I use it more as a sounding board, an editor and a review tool rather than a tool to write episode scripts.
[00:16:22] Now, when I've finished the draft of an episode, I’ll typically wait at least 24 hours, preferably more like two or three days, and then I’ll return to the script. I’ll review it myself, spend a couple more hours making changes, and I’ll also paste the entire thing into two or three different AI tools and ask them to do everything from fact-checking to grammar checking, and to give me feedback on the narrative arc, things that I’ve missed, and so on.
[00:16:55] I don’t always act on the feedback, but it certainly is useful to have it there as a second or even third opinion, albeit a non-human one.
[00:17:08] And then, once I am happy with it, I’ll press record.
[00:17:12] I almost always find something I’m unhappy with during the recording process, so I have to stop the recording and rewrite that bit of the script before re-recording it, but once that is done, my work with the script is over, and it’s passed to our audio editor before being passed to our in-house teacher who creates all the in-line vocabulary definitions, and the accompanying study pack.
[00:17:39] And…that, I guess, is it.
[00:17:44] It isn’t a huge secret or special sauce, but rather the product of doing something more than 500 times.
[00:17:53] Still, there are many times when I listen to something again and think “eugh, I should have said that in a different way, or it would have been interesting to have talked about x instead of y”.
[00:18:06] I try not to listen to the episodes again, because I know that if I did this, I’d be constantly worrying about going back to change certain things, and that would mean that there wasn’t any time for making new ones.
[00:18:22] And I also now don’t really look too much at how many people listened to certain episodes, especially not on YouTube. As you probably know, the algorithms on tools like YouTube and Spotify tend to gravitate towards clickbaity and extreme titles and subjects, so that’s why you see people making episodes about things like “10 English Words You Have To Know Today!” or “This Is Why You Can’t Speak English”.
[00:18:52] I know that I don’t want to make episodes about those sorts of things, and if I did, it would attract the wrong types of people to the show.
[00:19:02] What’s more, I often receive emails from people telling me how much they enjoyed episodes that didn’t do nearly as well in terms of downloads, so I now pay very little attention to the download numbers.
[00:19:16] What I do pay attention to, however, is what you, the wonderful listeners of this show and members of Leonardo English, tell me. So let me finish this slightly meta episode by sharing some interesting facts from this big survey that hundreds of you were so kind to fill out at the end of July.
[00:19:38] So, the most popular countries in terms of listeners are Spain, China, Brazil, Italy and Germany.
[00:19:47] Most listeners are over the age of 35.
[00:19:51] 69% of you say you listen for pleasure, while 47% list work as a reason.
[00:20:00] 75% of you also read books in English, and 66% of you watch TV or films in English.
[00:20:09] Only 12% of you go to an English-language school.
[00:20:14] 71% of you listen to other English podcasts–well done–and 65% of you say your biggest challenge is speaking fluently, no surprise there.
[00:20:27] I asked for feedback on how we could improve the show.
[00:20:31] There were some wonderful suggestions, but quite a few examples of the wide variety of people who listen to this podcast, and how you can’t please everyone.
[00:20:42] A bunch of you said you wanted more episodes about British society and culture, and about the same number of you said you wanted fewer.
[00:20:52] Almost as many of you said you wanted shorter episodes as said you wanted longer episodes, and there were as many of you who requested more complicated vocabulary as those of you who said you found it too hard to understand.
[00:21:07] We will, most probably, introduce some shorter episodes, and there are lots of more specific pieces of feedback that I’ll definitely be incorporating into the show, but these are just a few examples of how you can’t please everyone, and sometimes how you have to just do what feels right…
[00:21:28] And very interestingly, of those of you who are members, 70% said you use your membership either every day or several times a week, which just goes to show that it really is very useful to many of you.
[00:21:44] So, there we go, a peek behind the curtain, a look inside how these episodes are made.
[00:21:51] I get pretty regular emails from people saying that they are thinking of starting a similar type of podcast in Spanish or French or Italian, and asking me for “the secret sauce” of how to do it. As you can see, there really isn’t a secret, but it’s the product of doing something pretty much every day for five years.
[00:22:14] If you are thinking of starting a podcast like this in your own language — whether it’s Spanish, French, Chinese, whatever — my biggest piece of advice is just to start. Don’t worry if the first episodes aren’t perfect. As you now know, mine certainly weren’t.
[00:22:31] And as a final thought, thank you for listening, and an especially big thank you to those of you who are members or who have ever bought a course or one of our products.
[00:22:42] This show literally wouldn’t be possible without you.
[00:22:47] OK then, that’s it for this meta episode on how I make podcast episodes. I hope it was an interesting one, that you learned something new, and don’t worry, regular service will resume from the next episode.
[00:23:00] You've been listening to English Learning for Curious Minds by Leonardo English.
[00:23:05] I'm Alastair Budge, you stay safe, and I'll catch you in the next episode.
[00:00:04] Hello, hello, hello, and welcome to English Learning for Curious Minds, by Leonardo English, the show where you can listen to fascinating stories and learn weird and wonderful things about the world at the same time as improving your English.
[00:00:21] I'm Alastair Budge, and today we have an unusual type of episode.
[00:00:27] I’m not going to talk about a person’s life, a historical event, or some other weird and wonderful thing going on in the world.
[00:00:35] Today, I’m going to talk to you about how I choose, research, write and produce podcast episodes, and how this process has changed over the past almost six years of doing so.
[00:00:49] I’ve had quite a few requests for this type of episode, and a few weeks ago, I got it again from an interesting Brazilian member called Itamar.
[00:00:58] So, Itamar, and to everyone else listening to this, I hope you’ll enjoy this slightly meta type of episode. I’m also going to share a bunch of things I found out about the kind of people who listen to this podcast from a recent survey, so I think you might find that interesting, too.
[00:01:16] OK then, let’s get right into it.
[00:01:20] Depending on the way you calculate it, there are almost 5 million different podcasts.
[00:01:27] Not individual podcast episodes; 5 million podcasts, 5 million different shows.
[00:01:35] As I’m sure you know, there are podcasts about everything: comedy, TV shows, business, football, knitting, fishing, obscure historical periods, and, lest I forget, podcasts that help you learn a language.
[00:01:52] English Learning for Curious Minds–this podcast–is one of those.
[00:01:57] Of course, it isn’t the only one. There are tonnes of different podcasts that help you learn English, and they tend to follow similar formats.
[00:02:08] Some are conversation-based, with two teachers having a conversation about anything from their personal lives to a particular element of English grammar or language.
[00:02:20] Others might be one person giving motivational advice, or giving what’s essentially an English lesson in an audio format.
[00:02:30] This one is a little different.
[00:02:34] My approach right from the start has always been to present an interesting and thought-provoking subject in a way that allows you to learn English through listening to the episode, and allows you to learn weird and wonderful things about the world at the same time as you improve your English.
[00:02:56] This is now episode 565, so I have had quite a bit of practice making them.
[00:03:03] And in this meta episode, I’ll tell you what I’ve learned and what the process of making an episode looks like now.
[00:03:13] So, let’s go back to the start.
[00:03:16] When I first started, almost six years ago now, I had absolutely no idea what I was doing.
[00:03:24] I had quite a good idea of what I wanted the finished product to look like–a 15- to 20-minute episode that was both interesting and helped people improve their English–but I didn’t have much of an idea about how to actually make one.
[00:03:41] So…I just started, and figured I’d make it up as I went along.
[00:03:48] I chose some topics that I thought might be interesting, I did a bunch of research into them, and I pulled up Google Docs and just started typing.
[00:03:59] When I thought they were good enough, I recorded them using a microphone I borrowed from a friend.
[00:04:06] Then I listened to them again and…I thought they sounded awful.
[00:04:12] Firstly, like most people, I really didn’t like the sound of my own voice.
[00:04:18] And secondly, I thought the content of the episodes themselves sounded dull. I wouldn’t want to listen to them, so why should anyone else?
[00:04:29] In fact, I never released these early episodes. They were on topics I now realise were pretty difficult to make interesting in an audio-only format: there was one on how black holes work, and another one on how the international bond market works.
[00:04:48] Interesting and important subjects, sure, but quite hard to turn into an interesting topic for someone at B1 or B2 level.
[00:04:58] So, I scrapped them - they simply never saw the light of day.
[00:05:04] I decided to choose some slightly easier subjects to tackle, and things that I was more familiar with, and got cracking.
[00:05:14] That was in December 2019, coming on 6 years ago now, and I have just kept going, with a few changes and improvements along the way.
[00:05:25] So, how does it work now?
[00:05:29] Well, first comes the idea for the topic.
[00:05:33] If you are a seasoned listener to this show, you’ll know that we’ve covered a lot of different subjects.
[00:05:40] There’s been everything from episodes about the life of disappeared English aristocrats through to stories about the Vikings, how recycling works, conspiracies from ancient times, the list goes on.
[00:05:54] There are only really three things that need to be true for me to decide that they are worthy of an episode.
[00:06:02] Firstly, I need to find the subject interesting. Whether that means it’s an interesting story about a period of history, about someone who lived a particularly interesting or important life, or about a particularly interesting concept, it needs to be interesting to me.
[00:06:23] Secondly, it shouldn’t be about current affairs or a topic that will become irrelevant in weeks or months to come. My hope is that anyone can go through our library of episodes and choose one that we might have made several years ago, but it will be just as relevant today as it was when it was made. So something like King Henry VIII or Cleopatra - realistically, that isn’t going to change much, so it makes for a good candidate.
[00:06:57] And thirdly, it shouldn’t be about a highly controversial or divisive topic. I try to be as apolitical as possible in this show, within reason, and there are some topics that I know with complete certainty would end up flooding my inbox with angry messages on both sides of the argument. I’m sure you can imagine what might fit into that category.
[00:07:23] There are a few other things I take into consideration too, like whether a story can be properly told in an audio-only format. Some, like if you are talking about works of art or geographical borders, just don’t work quite so well without images, so I’ll also take that into consideration, and that’s the reason there aren’t many episodes about art.
[00:07:49] We also now make one three-part mini-series every month.
[00:07:54] Sometimes these will be about three different subjects that are all on a similar theme, but sometimes there will be a story that I just don’t think can be properly told in a single 20-minute episode, like the one on the Cambridge Five spy ring. This started as a standalone, single episode, but I ended up splitting it into three different episodes, a three-part mini-series.
[00:08:23] So, once a topic is chosen, what next?
[00:08:27] Well, I’ll usually try to get as much of an understanding of the story or the topic as possible before writing anything down.
[00:08:37] This might start with something as basic as skimming over a Wikipedia entry to try to get an overview of the topic, before going into more detail about particular events or parts of the story, reading opinion pieces, downloading datasets and trying to make sense of them, reading books on the topic, it can be anything that helps me understand the topic better.
[00:09:01] Sometimes I start from a position of relative knowledge, if I’m writing about something I already know about. I started this show while living in Malta. I lived there for seven years and made several episodes about things related to Malta, so I was able to write most of those without too much prior research.
[00:09:24] And I’ll often make episodes about people or historical events I’m already reading about, so I’ll be reading a book for fun, taking notes, and then when it comes to writing the episode of English Learning for Curious Minds, I just refer back to my notes, and the process is a whole lot simpler.
[00:09:45] Once I feel relatively comfortable with my own knowledge of the subject, I’ll make a rough plan and start writing.
[00:09:54] A rough plan for me is a very simple collection of five or so bullet points, with general ideas of how I want the episode to develop. Sometimes it’s easy, like if it’s about the story of someone’s life, which typically follows a chronological order, but sometimes it’s harder, and it takes a bit more time and thought to get it right.
[00:10:20] Then, I start writing.
[00:10:23] When I first started, I would write out general bullet points rather than the entire script, but when it came to recording, I found this meant I spent more time pausing and repeating certain parts in a different way, so now I write out the entire script, and I simply read it out loud.
[00:10:46] But I write as I speak, or at least I try to, so my hope is for an episode to sound more like a conversation, more like natural speech, than a formal written article read aloud.
[00:11:02] As you can probably tell, I use a range of different vocabulary and expressions. There is no particular grand science to this, but I try to use language that someone at a B1 level could understand at a bit of a stretch, language that someone at a C1 level could understand almost entirely, but that anyone at B1 to C1 level will learn from.
[00:11:31] This means not artificially dumbing down my language or using simple terms, but I might say something twice in two different ways — first naturally, and then in another, simpler way that will help you understand if you didn’t the first time.
[00:11:52] I speak more clearly than I would if I were speaking to a native speaker, and of course, I make sure that all of the more complicated words and phrases are clearly defined in our transcripts and study packs.
[00:12:06] There are around 100 of these advanced words and phrases per episode, so that’s more than 50,000 different definitions created, and when you remove the duplicates, the ones that have appeared twice or more, there’s something like 20,000 B2 and above words that you’d learn if you listened to every episode.
[00:12:29] Now, in terms of length, I aim to make each episode around 20 minutes long, with some coming in slightly under that and others a little over, which means the script should be between 2,500 and 3,000 words long.
[00:12:49] I try to write the script in one go, and this writing process typically takes anywhere from two to five hours. It’s rarely less than two, sometimes it can be much more than 5, but it tends to fall into that two-to-five-hour bracket.
[00:13:09] Now, you might think this is quick, or you might think it’s a long time, but all I’ll say is that it used to take me a lot longer. After doing it 565 times and writing well over a million words, I've had quite a lot of practice, so it would be strange if I hadn’t got quicker.
[00:13:31] And this brings me on to another important point.
[00:13:35] At various points over the years, I have tried a bunch of things to speed up the process.
[00:13:42] I've experimented with hiring external scriptwriters to write episodes of English Learning for Curious Minds. I posted a job for this once and got more than 500 applicants, and I trialled 30 different ones.
[00:13:58] Unfortunately–or perhaps fortunately–I never managed to find someone who was able to write to a level of quality that I was happy with, and that I felt was true to my voice and style, the voice and style of this podcast.
[00:14:14] In every case, when I would review the script that they had written, I found myself making so many edits and changes to their drafts, and starting to record an episode and realising that something they had written didn’t sound like me, that I realised it would have been quicker for me to just write everything from scratch.
[00:14:37] So that is now what I do: no external scriptwriters, everything on this show is written by me, Alastair.
[00:14:46] I do, however, have another scriptwriting colleague, in the form of various AI models I use to help me at different points in the planning and writing process.
[00:14:58] Specifically, I primarily use ChatGPT, Claude and Grok, but I’ll sometimes use other tools as well for a second or third opinion on something.
[00:15:09] During the planning stage, I will often tell ChatGPT what my proposed outline is, and ask it to point out areas I hadn’t thought about.
[00:15:20] If I get writer’s block, and I’m just not sure where to go next in a script, I might paste in what I’ve written thus far and ask it to give me three different routes to take it next.
[00:15:34] I don’t always use what the AI suggests, and if I do use them, I often find myself having to rewrite much of what ChatGPT provided, but they can be useful for giving me a fresh idea and helping me get back on track.
[00:15:52] I have, by the way, tested multiple times trying to get ChatGPT to actually write episodes in my voice. It does a better job than most of the human scriptwriters I've previously worked with, but it still sounds far too artificial, too unlike me, so I use it more as a sounding board, an editor and a review tool rather than a tool to write episode scripts.
[00:16:22] Now, when I've finished the draft of an episode, I’ll typically wait at least 24 hours, preferably more like two or three days, and then I’ll return to the script. I’ll review it myself, spend a couple more hours making changes, and I’ll also paste the entire thing into two or three different AI tools and ask them to do everything from fact-checking to grammar checking, and to give me feedback on the narrative arc, things that I’ve missed, and so on.
[00:16:55] I don’t always act on the feedback, but it certainly is useful to have it there as a second or even third opinion, albeit a non-human one.
[00:17:08] And then, once I am happy with it, I’ll press record.
[00:17:12] I almost always find something I’m unhappy with during the recording process, so I have to stop the recording and rewrite that bit of the script before re-recording it, but once that is done, my work with the script is over, and it’s passed to our audio editor before being passed to our in-house teacher who creates all the in-line vocabulary definitions, and the accompanying study pack.
[00:17:39] And…that, I guess, is it.
[00:17:44] It isn’t a huge secret or special sauce, but rather the product of doing something more than 500 times.
[00:17:53] Still, there are many times when I listen to something again and think “eugh, I should have said that in a different way, or it would have been interesting to have talked about x instead of y”.
[00:18:06] I try not to listen to the episodes again, because I know that if I did this, I’d be constantly worrying about going back to change certain things, and that would mean that there wasn’t any time for making new ones.
[00:18:22] And I also now don’t really look too much at how many people listened to certain episodes, especially not on YouTube. As you probably know, the algorithms on tools like YouTube and Spotify tend to gravitate towards clickbaity and extreme titles and subjects, so that’s why you see people making episodes about things like “10 English Words You Have To Know Today!” or “This Is Why You Can’t Speak English”.
[00:18:52] I know that I don’t want to make episodes about those sorts of things, and if I did, it would attract the wrong types of people to the show.
[00:19:02] What’s more, I often receive emails from people telling me how much they enjoyed episodes that didn’t do nearly as well in terms of downloads, so I now pay very little attention to the download numbers.
[00:19:16] What I do pay attention to, however, is what you, the wonderful listeners of this show and members of Leonardo English, tell me. So let me finish this slightly meta episode by sharing some interesting facts from this big survey that hundreds of you were so kind to fill out at the end of July.
[00:19:38] So, the most popular countries in terms of listeners are Spain, China, Brazil, Italy and Germany.
[00:19:47] Most listeners are over the age of 35.
[00:19:51] 69% of you say you listen for pleasure, while 47% list work as a reason.
[00:20:00] 75% of you also read books in English, and 66% of you watch TV or films in English.
[00:20:09] Only 12% of you go to an English-language school.
[00:20:14] 71% of you listen to other English podcasts–well done–and 65% of you say your biggest challenge is speaking fluently, no surprise there.
[00:20:27] I asked for feedback on how we could improve the show.
[00:20:31] There were some wonderful suggestions, but quite a few examples of the wide variety of people who listen to this podcast, and how you can’t please everyone.
[00:20:42] A bunch of you said you wanted more episodes about British society and culture, and about the same number of you said you wanted fewer.
[00:20:52] Almost as many of you said you wanted shorter episodes as said you wanted longer episodes, and there were as many of you who requested more complicated vocabulary as those of you who said you found it too hard to understand.
[00:21:07] We will, most probably, introduce some shorter episodes, and there are lots of more specific pieces of feedback that I’ll definitely be incorporating into the show, but these are just a few examples of how you can’t please everyone, and sometimes how you have to just do what feels right…
[00:21:28] And very interestingly, of those of you who are members, 70% said you use your membership either every day or several times a week, which just goes to show that it really is very useful to many of you.
[00:21:44] So, there we go, a peek behind the curtain, a look inside how these episodes are made.
[00:21:51] I get pretty regular emails from people saying that they are thinking of starting a similar type of podcast in Spanish or French or Italian, and asking me for “the secret sauce” of how to do it. As you can see, there really isn’t a secret, but it’s the product of doing something pretty much every day for five years.
[00:22:14] If you are thinking of starting a podcast like this in your own language — whether it’s Spanish, French, Chinese, whatever — my biggest piece of advice is just to start. Don’t worry if the first episodes aren’t perfect. As you now know, mine certainly weren’t.
[00:22:31] And as a final thought, thank you for listening, and an especially big thank you to those of you who are members or who have ever bought a course or one of our products.
[00:22:42] This show literally wouldn’t be possible without you.
[00:22:47] OK then, that’s it for this meta episode on how I make podcast episodes. I hope it was an interesting one, that you learned something new, and don’t worry, regular service will resume from the next episode.
[00:23:00] You've been listening to English Learning for Curious Minds by Leonardo English.
[00:23:05] I'm Alastair Budge, you stay safe, and I'll catch you in the next episode.
[00:00:04] Hello, hello, hello, and welcome to English Learning for Curious Minds, by Leonardo English, the show where you can listen to fascinating stories and learn weird and wonderful things about the world at the same time as improving your English.
[00:00:21] I'm Alastair Budge, and today we have an unusual type of episode.
[00:00:27] I’m not going to talk about a person’s life, a historical event, or some other weird and wonderful thing going on in the world.
[00:00:35] Today, I’m going to talk to you about how I choose, research, write and produce podcast episodes, and how this process has changed over the past almost six years of doing so.
[00:00:49] I’ve had quite a few requests for this type of episode, and a few weeks ago, I got it again from an interesting Brazilian member called Itamar.
[00:00:58] So, Itamar, and to everyone else listening to this, I hope you’ll enjoy this slightly meta type of episode. I’m also going to share a bunch of things I found out about the kind of people who listen to this podcast from a recent survey, so I think you might find that interesting, too.
[00:01:16] OK then, let’s get right into it.
[00:01:20] Depending on the way you calculate it, there are almost 5 million different podcasts.
[00:01:27] Not individual podcast episodes; 5 million podcasts, 5 million different shows.
[00:01:35] As I’m sure you know, there are podcasts about everything: comedy, TV shows, business, football, knitting, fishing, obscure historical periods, and, lest I forget, podcasts that help you learn a language.
[00:01:52] English Learning for Curious Minds–this podcast–is one of those.
[00:01:57] Of course, it isn’t the only one. There are tonnes of different podcasts that help you learn English, and they tend to follow similar formats.
[00:02:08] Some are conversation-based, with two teachers having a conversation about anything from their personal lives to a particular element of English grammar or language.
[00:02:20] Others might be one person giving motivational advice, or giving what’s essentially an English lesson in an audio format.
[00:02:30] This one is a little different.
[00:02:34] My approach right from the start has always been to present an interesting and thought-provoking subject in a way that allows you to learn English through listening to the episode, and allows you to learn weird and wonderful things about the world at the same time as you improve your English.
[00:02:56] This is now episode 565, so I have had quite a bit of practice making them.
[00:03:03] And in this meta episode, I’ll tell you what I’ve learned and what the process of making an episode looks like now.
[00:03:13] So, let’s go back to the start.
[00:03:16] When I first started, almost six years ago now, I had absolutely no idea what I was doing.
[00:03:24] I had quite a good idea of what I wanted the finished product to look like–a 15- to 20-minute episode that was both interesting and helped people improve their English–but I didn’t have much of an idea about how to actually make one.
[00:03:41] So…I just started, and figured I’d make it up as I went along.
[00:03:48] I chose some topics that I thought might be interesting, I did a bunch of research into them, and I pulled up Google Docs and just started typing.
[00:03:59] When I thought they were good enough, I recorded them using a microphone I borrowed from a friend.
[00:04:06] Then I listened to them again and…I thought they sounded awful.
[00:04:12] Firstly, like most people, I really didn’t like the sound of my own voice.
[00:04:18] And secondly, I thought the content of the episodes themselves sounded dull. I wouldn’t want to listen to them, so why should anyone else?
[00:04:29] In fact, I never released these early episodes. They were on topics I now realise were pretty difficult to make interesting in an audio-only format: there was one on how black holes work, and another one on how the international bond market works.
[00:04:48] Interesting and important subjects, sure, but quite hard to turn into an interesting topic for someone at B1 or B2 level.
[00:04:58] So, I scrapped them - they simply never saw the light of day.
[00:05:04] I decided to choose some slightly easier subjects to tackle, and things that I was more familiar with, and got cracking.
[00:05:14] That was in December 2019, coming on 6 years ago now, and I have just kept going, with a few changes and improvements along the way.
[00:05:25] So, how does it work now?
[00:05:29] Well, first comes the idea for the topic.
[00:05:33] If you are a seasoned listener to this show, you’ll know that we’ve covered a lot of different subjects.
[00:05:40] There’s been everything from episodes about the life of disappeared English aristocrats through to stories about the Vikings, how recycling works, conspiracies from ancient times, the list goes on.
[00:05:54] There are only really three things that need to be true for me to decide that they are worthy of an episode.
[00:06:02] Firstly, I need to find the subject interesting. Whether that means it’s an interesting story about a period of history, about someone who lived a particularly interesting or important life, or about a particularly interesting concept, it needs to be interesting to me.
[00:06:23] Secondly, it shouldn’t be about current affairs or a topic that will become irrelevant in weeks or months to come. My hope is that anyone can go through our library of episodes and choose one that we might have made several years ago, but it will be just as relevant today as it was when it was made. So something like King Henry VIII or Cleopatra - realistically, that isn’t going to change much, so it makes for a good candidate.
[00:06:57] And thirdly, it shouldn’t be about a highly controversial or divisive topic. I try to be as apolitical as possible in this show, within reason, and there are some topics that I know with complete certainty would end up flooding my inbox with angry messages on both sides of the argument. I’m sure you can imagine what might fit into that category.
[00:07:23] There are a few other things I take into consideration too, like whether a story can be properly told in an audio-only format. Some, like if you are talking about works of art or geographical borders, just don’t work quite so well without images, so I’ll also take that into consideration, and that’s the reason there aren’t many episodes about art.
[00:07:49] We also now make one three-part mini-series every month.
[00:07:54] Sometimes these will be about three different subjects that are all on a similar theme, but sometimes there will be a story that I just don’t think can be properly told in a single 20-minute episode, like the one on the Cambridge Five spy ring. This started as a standalone, single episode, but I ended up splitting it into three different episodes, a three-part mini-series.
[00:08:23] So, once a topic is chosen, what next?
[00:08:27] Well, I’ll usually try to get as much of an understanding of the story or the topic as possible before writing anything down.
[00:08:37] This might start with something as basic as skimming over a Wikipedia entry to try to get an overview of the topic, before going into more detail about particular events or parts of the story, reading opinion pieces, downloading datasets and trying to make sense of them, reading books on the topic, it can be anything that helps me understand the topic better.
[00:09:01] Sometimes I start from a position of relative knowledge, if I’m writing about something I already know about. I started this show while living in Malta. I lived there for seven years and made several episodes about things related to Malta, so I was able to write most of those without too much prior research.
[00:09:24] And I’ll often make episodes about people or historical events I’m already reading about, so I’ll be reading a book for fun, taking notes, and then when it comes to writing the episode of English Learning for Curious Minds, I just refer back to my notes, and the process is a whole lot simpler.
[00:09:45] Once I feel relatively comfortable with my own knowledge of the subject, I’ll make a rough plan and start writing.
[00:09:54] A rough plan for me is a very simple collection of five or so bullet points, with general ideas of how I want the episode to develop. Sometimes it’s easy, like if it’s about the story of someone’s life, which typically follows a chronological order, but sometimes it’s harder, and it takes a bit more time and thought to get it right.
[00:10:20] Then, I start writing.
[00:10:23] When I first started, I would write out general bullet points rather than the entire script, but when it came to recording, I found this meant I spent more time pausing and repeating certain parts in a different way, so now I write out the entire script, and I simply read it out loud.
[00:10:46] But I write as I speak, or at least I try to, so my hope is for an episode to sound more like a conversation, more like natural speech, than a formal written article read aloud.
[00:11:02] As you can probably tell, I use a range of different vocabulary and expressions. There is no particular grand science to this, but I try to use language that someone at a B1 level could understand at a bit of a stretch, language that someone at a C1 level could understand almost entirely, but that anyone at B1 to C1 level will learn from.
[00:11:31] This means not artificially dumbing down my language or using simple terms, but I might say something twice in two different ways — first naturally, and then in another, simpler way that will help you understand if you didn’t the first time.
[00:11:52] I speak more clearly than I would if I were speaking to a native speaker, and of course, I make sure that all of the more complicated words and phrases are clearly defined in our transcripts and study packs.
[00:12:06] There are around 100 of these advanced words and phrases per episode, so that’s more than 50,000 different definitions created, and when you remove the duplicates, the ones that have appeared twice or more, there’s something like 20,000 B2 and above words that you’d learn if you listened to every episode.
[00:12:29] Now, in terms of length, I aim to make each episode around 20 minutes long, with some coming in slightly under that and others a little over, which means the script should be between 2,500 and 3,000 words long.
[00:12:49] I try to write the script in one go, and this writing process typically takes anywhere from two to five hours. It’s rarely less than two, sometimes it can be much more than 5, but it tends to fall into that two-to-five-hour bracket.
[00:13:09] Now, you might think this is quick, or you might think it’s a long time, but all I’ll say is that it used to take me a lot longer. After doing it 565 times and writing well over a million words, I've had quite a lot of practice, so it would be strange if I hadn’t got quicker.
[00:13:31] And this brings me on to another important point.
[00:13:35] At various points over the years, I have tried a bunch of things to speed up the process.
[00:13:42] I've experimented with hiring external scriptwriters to write episodes of English Learning for Curious Minds. I posted a job for this once and got more than 500 applicants, and I trialled 30 different ones.
[00:13:58] Unfortunately–or perhaps fortunately–I never managed to find someone who was able to write to a level of quality that I was happy with, and that I felt was true to my voice and style, the voice and style of this podcast.
[00:14:14] In every case, when I would review the script that they had written, I found myself making so many edits and changes to their drafts, and starting to record an episode and realising that something they had written didn’t sound like me, that I realised it would have been quicker for me to just write everything from scratch.
[00:14:37] So that is now what I do: no external scriptwriters, everything on this show is written by me, Alastair.
[00:14:46] I do, however, have another scriptwriting colleague, in the form of various AI models I use to help me at different points in the planning and writing process.
[00:14:58] Specifically, I primarily use ChatGPT, Claude and Grok, but I’ll sometimes use other tools as well for a second or third opinion on something.
[00:15:09] During the planning stage, I will often tell ChatGPT what my proposed outline is, and ask it to point out areas I hadn’t thought about.
[00:15:20] If I get writer’s block, and I’m just not sure where to go next in a script, I might paste in what I’ve written thus far and ask it to give me three different routes to take it next.
[00:15:34] I don’t always use what the AI suggests, and if I do use them, I often find myself having to rewrite much of what ChatGPT provided, but they can be useful for giving me a fresh idea and helping me get back on track.
[00:15:52] I have, by the way, tested multiple times trying to get ChatGPT to actually write episodes in my voice. It does a better job than most of the human scriptwriters I've previously worked with, but it still sounds far too artificial, too unlike me, so I use it more as a sounding board, an editor and a review tool rather than a tool to write episode scripts.
[00:16:22] Now, when I've finished the draft of an episode, I’ll typically wait at least 24 hours, preferably more like two or three days, and then I’ll return to the script. I’ll review it myself, spend a couple more hours making changes, and I’ll also paste the entire thing into two or three different AI tools and ask them to do everything from fact-checking to grammar checking, and to give me feedback on the narrative arc, things that I’ve missed, and so on.
[00:16:55] I don’t always act on the feedback, but it certainly is useful to have it there as a second or even third opinion, albeit a non-human one.
[00:17:08] And then, once I am happy with it, I’ll press record.
[00:17:12] I almost always find something I’m unhappy with during the recording process, so I have to stop the recording and rewrite that bit of the script before re-recording it, but once that is done, my work with the script is over, and it’s passed to our audio editor before being passed to our in-house teacher who creates all the in-line vocabulary definitions, and the accompanying study pack.
[00:17:39] And…that, I guess, is it.
[00:17:44] It isn’t a huge secret or special sauce, but rather the product of doing something more than 500 times.
[00:17:53] Still, there are many times when I listen to something again and think “eugh, I should have said that in a different way, or it would have been interesting to have talked about x instead of y”.
[00:18:06] I try not to listen to the episodes again, because I know that if I did this, I’d be constantly worrying about going back to change certain things, and that would mean that there wasn’t any time for making new ones.
[00:18:22] And I also now don’t really look too much at how many people listened to certain episodes, especially not on YouTube. As you probably know, the algorithms on tools like YouTube and Spotify tend to gravitate towards clickbaity and extreme titles and subjects, so that’s why you see people making episodes about things like “10 English Words You Have To Know Today!” or “This Is Why You Can’t Speak English”.
[00:18:52] I know that I don’t want to make episodes about those sorts of things, and if I did, it would attract the wrong types of people to the show.
[00:19:02] What’s more, I often receive emails from people telling me how much they enjoyed episodes that didn’t do nearly as well in terms of downloads, so I now pay very little attention to the download numbers.
[00:19:16] What I do pay attention to, however, is what you, the wonderful listeners of this show and members of Leonardo English, tell me. So let me finish this slightly meta episode by sharing some interesting facts from this big survey that hundreds of you were so kind to fill out at the end of July.
[00:19:38] So, the most popular countries in terms of listeners are Spain, China, Brazil, Italy and Germany.
[00:19:47] Most listeners are over the age of 35.
[00:19:51] 69% of you say you listen for pleasure, while 47% list work as a reason.
[00:20:00] 75% of you also read books in English, and 66% of you watch TV or films in English.
[00:20:09] Only 12% of you go to an English-language school.
[00:20:14] 71% of you listen to other English podcasts–well done–and 65% of you say your biggest challenge is speaking fluently, no surprise there.
[00:20:27] I asked for feedback on how we could improve the show.
[00:20:31] There were some wonderful suggestions, but quite a few examples of the wide variety of people who listen to this podcast, and how you can’t please everyone.
[00:20:42] A bunch of you said you wanted more episodes about British society and culture, and about the same number of you said you wanted fewer.
[00:20:52] Almost as many of you said you wanted shorter episodes as said you wanted longer episodes, and there were as many of you who requested more complicated vocabulary as those of you who said you found it too hard to understand.
[00:21:07] We will, most probably, introduce some shorter episodes, and there are lots of more specific pieces of feedback that I’ll definitely be incorporating into the show, but these are just a few examples of how you can’t please everyone, and sometimes how you have to just do what feels right…
[00:21:28] And very interestingly, of those of you who are members, 70% said you use your membership either every day or several times a week, which just goes to show that it really is very useful to many of you.
[00:21:44] So, there we go, a peek behind the curtain, a look inside how these episodes are made.
[00:21:51] I get pretty regular emails from people saying that they are thinking of starting a similar type of podcast in Spanish or French or Italian, and asking me for “the secret sauce” of how to do it. As you can see, there really isn’t a secret, but it’s the product of doing something pretty much every day for five years.
[00:22:14] If you are thinking of starting a podcast like this in your own language — whether it’s Spanish, French, Chinese, whatever — my biggest piece of advice is just to start. Don’t worry if the first episodes aren’t perfect. As you now know, mine certainly weren’t.
[00:22:31] And as a final thought, thank you for listening, and an especially big thank you to those of you who are members or who have ever bought a course or one of our products.
[00:22:42] This show literally wouldn’t be possible without you.
[00:22:47] OK then, that’s it for this meta episode on how I make podcast episodes. I hope it was an interesting one, that you learned something new, and don’t worry, regular service will resume from the next episode.
[00:23:00] You've been listening to English Learning for Curious Minds by Leonardo English.
[00:23:05] I'm Alastair Budge, you stay safe, and I'll catch you in the next episode.
