Everyone seems to default to Otter.ai as the gold standard for meeting transcription. I'm skeptical.
Used Consensus on a recent technical deep-dive call. The transcript was... fine. But "fine" doesn't justify a platform switch or a new subscription.
My question isn't just about which one got more words right on a clean audio file. That's table stakes.
I want to know about the real breakdowns:
* Handling thick accents or multiple people talking over each other.
* Accuracy with technical jargon, product names, or code syntax.
* The actual cost of errors. If Otter gets 95% and Consensus gets 94%, but that 1% difference is on key decision points or action items, then the raw percentage is meaningless.
Anyone done a real side-by-side with messy, real-world meeting audio? Not marketing demos. I'm talking about exporting the raw text and diffing the crucial parts. The migration cost of moving an entire team's workflow is high, and accuracy is the only lever that might justify it.
Your vendor is not your friend.
Totally feel you on the jargon point. My team uses a ton of internal acronyms and Otter sometimes turns them into real words, which is confusing later. "BRD" becomes "bird" a lot, ha.
Has anyone tried feeding a custom vocabulary list into either tool? Does that even help with the overlap/accents problem, or is it just for isolated terms?