DictaFlow

Nuance Dragon costs $699. I charge $7 a month. Six months in, here's what I've learned.

When people hear the price gap, they usually think I'm missing something. Enterprise support, accuracy, reliability. But Dragon was built 20 years ago, and the codebase shows it. The installer is 8GB. Setup takes an afternoon. Half the bad reviews are just people trying to get it running.

I spent a weekend with modern Whisper-based transcription and got better out-of-the-box accuracy. No training, no profile setup, no reading passages aloud for 20 minutes. That gap exists because Dragon was never really rebuilt, just patched.

The customers are also completely different. Dragon sells to hospital procurement teams and law firms with IT budgets. My users are individual doctors who are tired of dealing with IT, developers who want to talk into their IDE, people who just want to write faster. That person was never Dragon's priority.

The interaction model was the thing I got wrong at first. Every dictation tool defaults to always-on listening. You forget it's running, say something weird, and it ends up in your document. Hold-to-talk, where you press a key to start and release to stop, is so obviously better that I genuinely don't get why nobody made it standard. Once people try it, they never want to go back.

The feature that gets the most comments is Actually Override. While you're talking, if you misspeak, you say a keyword and DictaFlow rewinds and keeps going. No keyboard, no stopping. I built it because I needed it myself. Turns out a lot of people feel the same.

The hard part has nothing to do with the product. Getting someone to trust a $7 tool over a $699 one is genuinely weird. Cheaper just feels worse to people, even when you can show them it isn't. I'm still figuring out how to close that gap. Word of mouth helps more than anything else I've tried.

If you're going up against a legacy product, the technical side is winnable. Distribution is the part that will actually take time.