Interviews News

Inside Python Podcasting: Interview with Michael Kennedy

The Python community has long been hailed as inviting to newcomers. But it’s now huge, which can be daunting. Fortunately, we have some guides that help you sort through all the packages and happenings. Michael Kennedy has long been one of those guides, especially with his long-running, well-known Talk Python To Me podcast.

After almost two years and 94 episodes, the podcast has covered many of the most interesting and important aspects of the Python world: SQLAlchemy, PyLadies, teaching high schoolers, data science, business … and yes, PyCharm. The episodes have a fresh, conversational feel that manages to be both approachable to newcomers while useful to veterans. Moreso, you get to know and appreciate people in the community as much as the code they produce.

Recently, Michael has teamed up with Brian Okken of Test and Code Podcast fame to produce another podcast called Python Bytes which surveys the latest in Python, in brief episodes.

PyCharm caught up with Michael on all this Python podcasting activity in the interview that follows.

Hi Michael, thanks for doing the interview. Let’s start with a 2016 retrospective of all Python podcasting. Where are we and how did we get here?

You’re welcome, it’s great to be here.

If you compare the end of 2016 with the beginning of 2015, it’s night and day. Early 2015 was a pretty dark period for Python podcasts although it saw amazing growth of podcasts in general. 2015 started with exactly 0 active Python podcasts.

I was really keen to hear these stories and meet the people as you described in the introduction. Yet for Python podcasts, all we had available were a few excellent but older, on the verge of outdated, episodes from past podcasts. So I decided (in March 2015), that if I was to get the podcast I wanted, I’d have to take the chance and put myself out there and start one myself. That’s how Talk Python To Me came into existence.

In some crazy coincidence, Tobias Macey and Chris Patti, launched podcast.init within a week (neither of us were aware of the other).

Shortly thereafter, I met up with Brian Okken (he works near me in Portland, OR) and talked about podcasting. He launched Test and Code (largely focused on Python, especially in the beginning) a few months after the first two shows.

Around the same time, Partially Derivative (the data science podcast), took a strong turn towards Python as part of their focus. Since then we’ve had The Python Experience and Import This: A Podcast for Humans launch. So it’s a great time to engage with the community and listen to the podcasts around Python and related technologies.

That’s a lot of activity. What factors do you think are driving it?

That’s a good question. I don’t know the answer entirely but I see two forces at work:

  1. Python is exploding in popularity and this is largely due to it expanding into areas it was less popular previously (think data science over the past few years). This means the audience is growing, the desire for these shows is growing, and there are many newcomers who seriously benefit from the look behind the code that podcasts can give you.
  2. I think the success that Talk Python To Me and Podcast.__init__ have had is an inspiration to others. Seeing Talk Python do so well, gain such an audience (we are about to pass 3,000,000 downloads), and frankly appear to be fun to be part of what has encouraged people who were considering starting a podcast to actually take action.

What are the reasons and factors that motivated you to start podcasting?

I’ve always wanted to do a podcast since I first started listening to them back in 2004.

However, it wasn’t until 2015 that things that I felt were essential actually lined up: a gap in the podcasting space where I was both enthusiastic and qualified to fill and the motivation to take the step to actually do all the work to build the show, the website, the format, etc, etc.

Explain the formats of Talk Python to Me and Python Bytes, and what drove the creation of the latter.

I consider Talk Python to be the audio equivalent of long-form writing. The conversations are highly researched by both me and the guests. The topics dig deep into an area (many times there is a single topic discussed for an hour). It’s highly produced. I spend about 5-10 hours on research and prep for each episode and probably 5 hours after the recording before it goes live.

Python Bytes is, as the name would indicate, shorter and made of many pieces. It’s 6 topics covered in 15 minutes total (so approx 2-3 minutes each). If you want to just keep abreast of the Python news, new projects, and releases for the week. It’s a great format for this. Also, I guess Python Bytes is very timely but makes less sense to go back and listen to an episode from a year ago whereas Talk Python To Me is much more evergreen content.

The motivation to create Python Bytes was my desire to create a show where we could talk about timely issues but that was contrary to my goals for Talk Python.

What’s it been like having Brian as a collaborator and how does that change the process?

It’s been a real pleasure to work with Brian. It’s a totally different format than I’m used to.

Usually, I can let the guests talk and I just ask questions. With Python Bytes and two co-hosts, it’s much more on the two of us to carry the conversation. That said, we are getting good at it I think and the format is starting to flow.

How do you choose your topics? And what do you see as your target audience for these topics?

I get topics from many sources. Originally, I chose topics I was familiar with and already knew: MongoDB, SQLAlchemy, and Pyramid.

After 95 topics, we are well outside of my area of expertise. Now I just look for things that are catching on with the community or seem interesting. I love the topics where it’s Python + (a thing), like the topic on Python at CERN / LHC and Python in Astronomy.

I get a lot of topic recommendations from my listeners and I really appreciate that too.

The target audience is really two groups. I try to keep the shows impactful and not watered down. I want people who have been developers for 10 or 20 years to get real value from listening. Yet, I try to avoid acronym soup or just diving deep right away into a topic because it’s important to me that these episodes are really accessible to newcomers even if they don’t know everything, they get enough to keep listening and learning.

For full disclosure, we had you for a PyCharm webinar. I enjoy those, but it’s certainly some work. Talk Python is far more frequent and likely far more effort. Is it still fun?

Yes, it is a ton of work (like I said, just Talk Python is 15 hours a week and Python Bytes is probably another 3-5 hours). But it’s absolutely fun and I’m really honored to be able to do it. Honestly, I can’t believe I get to spend a large part of my professional life doing this kind of work.

If you think of dream-come-true jobs, the ability to spend your time researching cool projects, talking to the founders and people you look up to in the community and sharing that conversation with many thousands of people every week, well, that is right up there.

I’ve always been interested in the business side of Python. Tell us a little about your training and Kickstarter, with a reasonably honest assessment of whether the podcasting is helping.

I am very interested in the business of open source in general and Python in particular. I have had the opportunity to talk to a few people doing very cool things to build successful and thriving businesses on open source and Python. You can hear their story on these episodes:

  • #34, Continuum: Scientific Python and The Business of Open Source with Travis Oliphant
  • #50 Web scraping at scale with Scrapy and ScrapingHub with Pablo Hoffman
  • #59 SageMath – Open source is ready to compete in the classroom with William Stein

I am happy to talk about Talk Python To Me and the business side.

Quick bit of history: I started the podcast in early 2015. I went independent with the podcast as my only source of income to focus on it and on growing it. That was Feb 2016. In April 2016, I launched my online Python training company on Kickstarter.

I see the podcast and the training courses as a continuum on a spectrum rather than two distinct things.

When people want to learn about Python, dig deeper into the back story on a web framework or ORM, they can listen into Talk Python To Me and get inspired. When they are ready to move from inspiration to action and learn that framework, the training courses pick up where the podcast stops with hours of hands-on training and lots of "live" demos they can code along with.

I was willing to quit my job and focus on this path because the podcast was generating enough revenue to basically pay the important bills (mortgage, utilities, food). I had dreamed of launching an online training business and the podcast gave me the safety net to do this without jeopardizing my family. I actually discussed this in depth with John Sonmez on episode 71 of Talk Python To Me.

The kickstarters were more successful than I expected and really proved the idea of my style of classes. I did two kickstarters. One in April for my Python Jumpstart by Building 10 Apps course. The other in August for my Python for Entrepreneurs course I did in partnership with Matt Makai from Full Stack Python. They both raise about 18x what I was asking ($33k and $39k respectively).

These courses are a lot of work. But I love making them and the response has been super positive. My goal is to make 20 courses of the next few years. Fingers crossed.

What are some programming podcasts that you have listened to?

I love podcasts. Here are the tech-focused ones I love.


  • Partially Derivative
  • Test and Code
  • Podcast.__init__

The personalities of developers:

  • Away from the Keyboard
  • Developer on Fire

Amazing deep look at tech like solar energy space flight, etc.:

  • .NET Rocks’ Geek outs episodes (got to the tags section, chose these)


  • Risky Business

Business of software:

  • Exponent
  • Startup podcast

Looking at 2017 and podcasting, what’s one thing you think will or should stand out this time next year?

I think we’ll see the Python vs. Legacy Python divide essentially settled in Python’s favor (that is Python 3 vs. Python 2, reference). We’ll also see more use of the term Legacy Python.

In podcasting, I think we’ll see:

  • more Python podcasts created.
  • podcasts will continue to “grow up” where people put in more work than just hitting record and posting episodes (as we saw years ago). This is a continuum and we’ll move more towards polished shows.

Thanks again for joining us. As an exit, tell us about the podcasting booth at PyCon.

You’re welcome, thank you for the opportunity. Please come and hang out with the Python podcast community at our booth at PyCon in Portland this year. We’ll have Talk Python To Me, Python Bytes, Test and Code, Partially Derivative, and Podcast.init all there.

I’ll be giving away some of my course content. We’ll be doing live recordings and open sessions. I hope to meet many of you there.

image description