AI-driven transcription tips - the why and the ROI

Surveys consistently indicate 20-40% or more practitioner time is spent on administration - when it comes to file noting and composing documents, let technology do the work for you.

Start by assessing the % proportion of client conversations you have via video conference, face to face, and phone. Video conference is a good place to start as it’s often a simpler matter of applying a high performing transcription service to the recording option. Extra considerations:

Maximise audio quality

Data security - who is providing the services?

Ease of recording and transcription - seek integration

Pause and resume functionality

100% disclose you are recording at the beginning

Structure the formal part of your conversations with a clear sequence to next steps

Look for ease of editing the transcription, rapid advancements now available

Many transcription providers show confidence levels as one measure of performance (80-90% accuracy is reasonable with high audio quality)

Documentation - many of us can now type at over 120 words per minute at a high grammatical accuracy thanks to voice typing. These solutions (Microsoft Dictate and Google Voice Typing) are the best starting place to prove AI transcription performance to yourself, before moving to multi-speaker solutions. Smartphone dictation options are also under-utilised (OneNote and Google Keep).

Any significant process change should have a target ROI. With productivity initiatives, often best-estimate time savings (eg. hours per day / week) with a realistic $ per unit valuation are a good place to start when evaluating the additional spend. You want a practical outcome measurement approach with ongoing checkpoints (process completion and turnaround times are also good to track). Qualitative assessments are also important: focus on clients, speed of follow ups, greater detail in record keeping.