Transcribing Podcasts

Erick Schonfeld at the Business 2.0 blog found a company that is tackling one of the opportunities I mentioned in an earlier post – providing transcriptions of recordings to podcasters so that they can post searchable text along with the audio file.

CastingWords is not live yet, although they have a landing page up with some limited search capabilities and an email address for inquiries. Eric spoke with the company and gathered addtional information – apparently they are leveraging Amazon’s Mechanical Turk to produce human-created transcriptions (with software cleanup) of podcasts.

We’re a super early stage startup – the idea is to sell search ads against the keywords in the transcriptions, which we generate using Amazon’s Mechanical Turk ( and some whizzy software. But we need capital to pay all those MTurk workers, so in the next few weeks we’ll be opening a store. It will allow Podcasters to purchase transcriptions of their shows. We’ll do the transcriptions and give them a full transcription – not just chunks pulled back by the search engine if it happens to index their show. Of course they automatically get listed in the engine, so this arrangement should drive traffic to them, get them transcripts, and get us the cash needed to keep on transcribing…Link

This is a space I’ve been watching, as companies race to provide either human-created or automated solutions.

I was excited by a Wired article on November 30 that discussed two companies, Podzinger and blinkx, that are attempting to automatically scan and transcribe podcasts. But after a review of their services I came away disapointed and decided to wait before reviewing them. Neither are ready for prime time yet.