In spring 2020, my podcast app provided some 2 million podcasts to choose from... of which nearly half a million were in the 'science and technology' category, and more than 100,000 more covered 'government and politics'. So — I wondered, as I spent my 15-minute train ride to Brussels one day flicking through Apple's colossal podcast directory — why did not a single active podcast focus on science advice for policy?

Over the coming weeks, this innocent question gradually morphed into a far more dangerous one: What's stopping me?

I knew there was a potential audience out there, because the kinds of people who enjoy thinking about the science-policy interface were the exactly kinds of people I worked with every day. And everything I read told me that the barrier to entry for podcasting was quite low: apparently all you need is a good microphone, some editing software, and a bit of free time. (This last part turned out to be a bit misleading, but I didn't learn that until it was far too late.)

And, most dangerous of all... wasn't SAPEA the ideal organisation to create something like this? After all, alongside our central mission to provide evidence to the Commission as part of the Scientific Advice Mechanism, we had a mandate to promote debate in the wider community about the role of evidence in policymaking. And a podcast ought to be a great way to do that. So... if not us, then who?

Spring 2020 turned out to be both a brilliant and a terrible time to implement my new idea. As the first Covid lockdowns encircled the world, we had just cancelled all our outreach events and were suddenly desperate to find new ways to communicate with audiences. So a corner of my bedroom became a makeshift recording studio; I ordered a microphone, oiled my squeaky chair, and started firing off some emails to pull in some favours. Luckily, it wasn't like my guests had anything better to do.

On the other hand, everything I had read about the demographics of podcasting was turned upside down. You see, common wisdom pre-2020 was that podcasts occupy a strange liminal space between work and entertainment, between on-the-clock and off-the-clock activities. People listen to work-related podcasts, yes — but nobody sits down with their colleagues at 9:30 on a Monday morning and works through the latest playlist. Instead, podcast listening is invariably a solitary experience: episodes are downloaded to our brains in the strange early-morning dead time between home and office, when we are walking, cycling, riding the bus or looking for a parking space.

But of course, along with many other things, Covid killed the commute. Universal lockdown means no transition from home to work; no transition from home to work means no liminal space between the two; no liminal space means no podcasts. I was launching a new communications channel at a time when quite possibly nobody in the world would be interested in engaging with it. Or so I thought.

In fact, the opposite was true. I began to read about how Covid had ushered in a new golden era for podcasting. Forced to merge their personal and professional lives, and with most social activities on indefinite hold, people started consuming audio media all the time: doing the ironing, taking a lunch break, taking their state-approved 30-minute walk round the local neighbourhood, falling asleep at night. And I saw the same phenomenon with our new venture: our first episode found 60 listeners (brave souls!), our second approached 100, our third around 180, and so it grew, and grew.

More than two years later, the Science for Policy podcast has delivered a new episode every fortnight, interviewing around 100 guests in total from the worlds of politics, science advice and academia. Our loyal audience is basically science for policy geeks: people who work on either side of the science-policy interface, who study it, or who aspire to be part of it. We have featured politicians, philosophers, scientists, students, civil servants — and of course current and former science advisors from East and West Europe, North and South America, Australasia and Antarctica. And the whole thing is still produced in-house by SAPEA, largely from the same corner of the same bedroom.

Of course, as with any exploratory project, some things are easier than expected and some things are harder. Here are ten lessons I’ve learned from the first two years.

1. People like talking about themselves. I set my sights high from the start and targeted the top names, but I've still been amazed that nearly everyone I’ve approached has enthusiastically accepted the invitation. Two years later, there is absolutely no shortage of interesting, articulate guests who are happy to make themselves available. Now the podcast has enough street cred in the community that pretty much everyone I invite has already heard of it, and to be honest, I get the impression that some people are just waiting to be asked!
2. Niche is everything. We stumbled on an untapped need, and as a result our growth in listeners has been both sustained and entirely organic. I’ve never had to spend a cent on advertising, because the biggest evangelists are our existing listeners and guests. It helps that we are basically not competing with anyone else: INGSA created its own podcast a couple of years after us, but it's very different in style and scope.
3. The learning curve is steep. Most commercial podcasts have at least a producer who plans and sets up episodes, a presenter who researches and asks the questions, and an engineer who edits the resulting audio. Most small businesses outsource the job, and now I know why: I definitely underestimated the skillset needed to do all three! The time commitment is not so daunting — it averages out at about 8 hours per 45-minute episode — but it was a big task to learn audio editing and interviewing skills from scratch at the same time as continuing to do my day job. (Listen to the first few episodes and you'll hear just how bad I used to be!) Thankfully, as our team has grown, I now have quite a bit of support from colleagues. And the Board of SAPEA — my bosses — has been patient and supportive. Of course, it helps that the project has been a success...
4. 4. Europe and the US are different. About half of our listeners are spread across Europe, and the other half are global, largely the US, Australia and New Zealand. I've learned over time that these audiences are very different. In the US, the audience is very diverse, including many senior academics and political staff. In Europe, it skews more heavily towards younger people, especially early-career academics. There is a big overlap in interests, thankfully, but it's been fun to see how different episodes have been more or less popular in different places around the world.
5. Variety is the spice of life. Initially, I wanted the podcast to be precisely focused on scientific advice to policy. Over time, though, the topics have become a little looser, and with hindsight I think it's better that way. After all, if you work at the science-policy interface, you're probably also at least a bit interested in policy synthesis within government, public trust in expertise, fighting disinformation, or the role of pop psychology in politicians' decision-making. Plus, after more than 2 years, we have covered lots of the "core topics" pretty thoroughly, so it's natural to want to branch out a little. Admittedly, one or two episodes have slipped through the net that are really a bit too far from the core topic, but lessons have been learned!
6. The pull of the UK is strong. There are a few reasons why I always seem to end up with more Brits on the podcast than any other nationality. The fact that it's an English language show is definitely part of it, and probably my own connections as a former Brit help too. But the biggest cause seems to be simply that there is a lot of expertise — in both UK universities and the UK government — on topics related to science and policy. It's a constant challenge to maintain a reasonable geographic spread of guests.
7. Issues, not institutions. One point I'm constantly having to explain to prospective guests is that the podcast is not a place to talk about their institutions, or indeed themselves — it is a place to talk about ideas and issues. Academics tend to be pretty comfortable with this, as ideas are their stock-in-trade, but civil servants and politicians tend to find it rather more surprising. My reasoning is simple, though: our listeners spend their free time with us because they want to think about interesting topics, and if they only heard people advertising their own work, most of them would switch off very quickly.
8. Podcasts operate in the dark... It's a strangely one-sided medium. You put your episodes out there, and you hope people will listen. There are no debates or comments, no feedback from the audience, at least initially, except bare listener numbers. And to make matters worse, those numbers aren't reliable either, because each podcast directory (Apple, Spotify, Google, YouTube, and about 20 smaller ones you've never heard of) reports slightly different data — some record downloads, some listeners, some subscriptions, some plays — so my estimate of about 1000 listeners per month right now is partly based on guesswork too.
9. ...but that makes unsolicited feedback all the more rewarding. When the audience does get in touch, it's lovely. Sometimes listeners write to say what they thought of a recent episode (good or bad!). Sometimes they leave reviews on apps or comment on videos. Fairly commonly these days, when we invite a guest to appear on the show, their first response is "Oh wow, I've been a listener for months!". I know of at least two academic collaborations which have started because of connections made through the podcast. It's nice to know that it's making a difference to the community here and there.
10. It's fun to have fun. I take the weird not-quite-work, not-quite-leisure positioning of podcasts to give me licence to play around sometimes, and that keeps me sane in the otherwise quite strait-laced Brussels bubble. I often ask guests about their hobbies or try to find something weird in their biography to bring up. And one of my favourite episodes was a full-on interview with the science consultant for Star Trek, who brought up many fun parallels between advising TV producers and advising policymakers — as well as explaining why she personally would not step into Kirk's transporter beam.

In conclusion, I'm often asked by science communicators if I'd recommend starting a podcast. It's a hard question to answer. Frankly, it's a lot of work, and in an already overcrowded market, any podcast that just aims to promote its creator will rarely have much traction. People don't listen to marketing material for fun. On the other hand, I'd say if you have an idea for good content and an audience that will lap it up, and you either have time to dedicate to learning new skills or money to pay someone else who has them — then it can be a impactful, fun, and very rewarding project. Drop me a line if you decide to go for it!