image When people podcast today, they rarely think about some of the issues that this new medium brings. When I think of podcasting (or blogcasting if you're trying to be different) I think of the following issues that you too should consider.

Discoverability - When creating a podcast, people will typically post something as part of a series, and then hope that people find out about the series through searching its title or description. That's an okay model if you're regularly posting podcasts, but what if you're just featuring an important interview or documenting something via audio that you don't plan to do regularly? How will people find this content today?

Machine-created transcription is really hard. At Microsoft, we're looking at this problem several different ways, including speech indexing. Think of this as a search engine that can index spoken words. We have this technology already in use today as part of OneNote 2007, and we're always thinking about other ways to leverage this technology.

You could also consider tagging your own material (or waiting for the community to do it for you), but that takes time and a considerable audience to appropriately tag your content for future search. While that is beneficial, it's hard to scale without a lot of taggers, and it's really one-sided if you try to do it yourself (since tagging benefits from the power of crowds to tag content).

Accessibility - What are you going to do if a visitor to your site is deaf or blind? How are they supposed to consume content if it's only available in audio or audio/video form? Text transcription is an option, but it's really expensive if you do it right. Machine transcription isn't very accurate either. When creating a podcast, think about where else this content might be. Are there components of this content available in other mediums that can be accessed by the visually and hearing impaired?

Localization - If your content is spoken or displayed in any language, you risk alienating any other viewers that don't speak that language. Web page translation is still pretty poor today, but continually improves. We have a Windows Live Translator service that can translate entire web pages or bodies of text. The service is very similar to Google's language tools but the Live Translator has the added ability to try to not translate computer-related statements (like code).

Anyway, text is pretty well translated today, but how do you do the same for your audio/video? Providing localized transcripts or voice over is really expensive, and doesn't scale well if you plan to do this often. Think about how you might affect your audience if they don't speak English (or the language you're recording your audio in).

Editing of completed content - Once you've published an audio or video podcast, what happens when the content is out of date? Do you treat it like a blog and not bother to update it? Do you need to keep your source files around indefinitely thinking that you might update the content at some point? If keeping this content updated with the latest information is important, then you'll want to really think about how to edit your content and maintain its accuracy over time.

Now with all that said, I don't think these are reasons why you shouldn't create podcasts. Instead, think about these topics as you podcast, and determine if any or all of them might change how you'd produce, edit and distribute your content to your readers/viewers/listeners. These issues are things to keep in the back of your mind, but by all means - get out there and create new and interesting content as many ways as you can.

Have you tried podcasting? Share your story by posting a comment in this entry.

Hunter Donald
Program Manager - Knowledge Management Strategy
Commercial Technical Support