Here's an interesting case study that has an unusual spin on the digital dictation versus speech recognition debate...

In mid-2008, Queensland (Australia) based law firm Payne Butler Lang began to look at their dictation processes and recognized the need for a more robust, reliable product to replace their current, aging digital dictation solution implemented four years prior. According to practice manager Malcolm Sinclair, the objectives and requirements were clear. “We were already using a digital dictation system within the firm, but it was very old fashioned technology that we were beginning to outgrow, and the product really did not have the back-up facilities we required. The idea of a digital dictation product with speech recognition capabilities really appealed to us as it would allow us to adequately improve efficiencies within the firm.” But, during the procurement process, the firm spoke to a number of dictation suppliers and discovered that there weren’t many suppliers out there on the market that had the speech recognition component.

After a successful pilot period that involved the managing partner and his practice administrator, the Winscribe Dictation with speech recognition was rolled out firm-wide, user by user, with authors dictating using a mixture of Olympus DS 4000 and 5000 digital handheld recorders.

Since implementing Winscribe Dictation with speech recognition in late 2008, the firm has seen significant improvements on their previous system. “We conducted some in-house testing that involved secretaries typing dictations with the old system, versus using Winscribe plus speech recognition, and we were surprised by the results. Typing was 50% faster using WinScribe with speech recognition. From a productivity perspective, our managing partner is currently getting around 98% accuracy with the system. Even when we have new people in the firm start using it, they always start off with around 80% and rapidly improve from there as training on the voice profile is improved. Our system is set up in quite a unique way to allow authors to use templates. For example, one of our managing partners currently utilizes pre-form letters through Winscribe’s speech recognition component, whereby he speaks the name of a particular document that he would like to dictate into (he currently has a list of around 80 forms pre-loaded onto the system to choose from); and the document template appears on his screen ready to dictate. This means fewer corrections, when the secretary receives the document and faster output.”