PodCraft | How to Podcast & Craft a Fantastic Show Research as a Creative Partner, With Tom Webster of Sounds Profitable
Have you ever thought about publishing a report in your niche? A “state of the industry” piece that delivers fresh data, insights, and analysis to your audience.
Original research can position you as an authority while sparking conversation, attracting media coverage, and opening new opportunities.
But gathering and reporting data is both art and science, and common mistakes can quickly undermine your work.
Fortunately, we’re joined by Tom Webster of Sounds Profitable, one of the most trusted voices in podcast research.
With more than 30 years of experience, Tom shares practical advice on audience research and presenting original data. Whether you are planning a full industry report or simply running an audience survey, this conversation will help you avoid common pitfalls and prepare for success before writing your first question.
Here are some of Tom’s key insights from our chat:
Research Fuels Creativity
“It’s always bothered me that people viewed the research and data side of things as not creative, when actually what it gives you is constraints. And constraints are really the key ultimately to creativity.”
Far from being dry or restrictive, data gives you the boundaries that spark innovative thinking. By knowing how your audience reacts, you can sharpen your message, test new angles, and create with greater confidence.
Ask Better Questions
“Writing a question is one of those things that everybody thinks they can do and almost nobody does it well, because it requires a really bizarre way of thinking.”
Good research starts with good questions, but survey design is a skill in itself. A poorly phrased question can confuse your audience or skew your results beyond usefulness.
“Don’t start writing any questions until you’ve had conversations with listeners. Those chats don’t give you the final answers, but they show you what you should ask about.”
Begin with real conversations - virtual coffees, quick calls, or informal chats. Use your audience’s own words to shape your survey options, ensuring the language resonates with them rather than sounding like a form filled out by a stranger.
“If you ask a question and you don’t know what you would do with the result of any of the answers, don’t ask the question. Don’t waste people’s time.”
This is the ultimate filter. Every question in your survey should serve a purpose. If you don’t know how you’ll act on the response, cut it.
Research to Know vs. Research to Show
"Broadly, there’s two kinds of research in anything. There’s research to show and there’s research to know. And I like to specialize in research to know."
So what's the difference?
“I would often be asked by people, I want to do a survey that shows this. I want to do a survey that shows that this product approach that our company uses is actually the best. That’s research to show. And I always tell people, be prepared not to get the answers you like.”
True authority comes from being curious, not from trying to validate a pre-set agenda. If your findings surprise you, lean into that.
"If the research comes back credible, without obvious flaws, and it contradicts your original hypothesis, the best thing you can do is document it honestly. Share the story: explain what you expected, why you thought the outcome would be different, and then walk people through the actual findings. Reflect on what surprised you and what might have made the difference. That's the essence of thought leadership."
Audiences and peers will respect you far more if you publish results that challenge assumptions, even your own.
“…if you can’t tell a story with a particular data slide, then don’t include it. And that’s not necessarily cherry picking, that’s just this did not come back as an interesting finding.”
Not every data point belongs in your final report. Great reporting is about focus: highlight what tells a meaningful story, and don’t overload your audience with filler.
Be Transparent
“The magic word is respondents. You can’t say ‘audiences say this,’ but you can say ‘53% of respondents said this.’ You’re never going to go wrong there.”
Every survey has its limitations. Maybe your responses came from a mailing list, or from people who clicked a link in your podcast notes. That’s fine - just be clear about it. Transparency builds trust, while over-claiming erodes credibility.
"You don't have to denigrate your approach. You don’t have to talk yourself down. I’m just a big fan of just being very clear about what you did. Just tell them what you did."
A simple "Methodology" paragraph in your report will do the trick. No need to get granular with the details, but what were the places, platforms or methods you focused on to collect responses?
“…if you have a study that has 500 respondents, I think it’s just fine to say men say this and women say that. I think it’s probably just fine to say that 55 plus says this and 18 to 34 says that. But actually look at the number of men 18 to 34 in your study… you’re in the low double digits, right? And that’s where you want to be very, very careful.”
In other words, don’t slice your sample so thin that the numbers stop being meaningful. Sometimes it’s better to give raw counts than percentages when subgroups are small.
And... Iterate!
“…one of the things that it’s really difficult to do in a single survey is report a correlation… I think what you can say is this sample did this and this. And here’s the key, if you’re a curious person and you want to get better, is you iterate. The next time you do a survey, you ask about that correlation specifically, you make it specific and then you see, okay, that hypothesis was right. It’s a scientific method.”
Don’t try to force causation out of one dataset. Treat each survey as a stepping stone in a longer journey of discovery. If you can build on your data, you'll begin to spot interesting patterns and trends.
A huge thanks to Tom for sharing his insights and experience. SoundsProfitable.com is the main hub for keeping up with his work in the podcasting space!