Could voice recognition software save financial advisers hours every week?
We could all do with a little more time in our days — could voice recognition software help you get just that? In his latest article for Aviva, Ian McKenna, Managing Director of FTRC, shares his advice for making the most of this technology.
If there is one commodity anyone in business always wants more of, it is time. In my experience if you ask any IFA what they would like to do more of the answer is invariably spend time advising clients.
So if the technology is available that could liberate large amounts of time you might expect it would be widely embraced, especially as it can be used from any smart phone. In this article I would like to look at the various types of Voice Recognition software which can save advisers many hours every week.
A brief history of dictation
40 years ago most people working in financial advice spent large amounts of time dictating letters and reports which would be transcribed so they could be sent to clients and business partners.
Two generations later much of this work is carried out by advisers and paraplanners themselves, but how many of us have ever had any formal typing training? We have moved from a world where an audio typist could produce perhaps 100 to 120 words per minute to one where highly qualified professionals struggle to achieve a third of these speeds.
Such inefficiency is however totally unnecessary. Voice recognition software, which can convert every day speech into written words on smart phones, tablets and computers is widely available and indeed has been for a decade or so. Most people actually talk at around 300 words per minute so voice recognition is far more efficient than even the best audio typist could achieve.
Even if you halve this speed to allow for proofing the text this is still going to be three times more efficient than most people can achieve typing themselves. If an adviser spends three hours a day on reports and correspondence, even assuming a 50 word per minute typing speed, which is probably generous, voice recognition has the potential to liberate two of those three hours a day. What could you do with an extra 2 hours a day?
Voice recognition technology is actually nothing new. The second technology column I ever wrote back in 1995 was about one of the early versions of such software. In those days it took considerable patience and training but today these products can be used instantly with no training at all.
You already own the software
The software sits on all smart phones it’s just that most people don’t realise it’s there. It is an embedded part of both the Apple iOS and Google Android operating systems.
If you are an Apple iOS user voice recognition can be accessed from a microphone icon on the standard keyboard. If this does not appear go to Settings> General> Keyboards and turn on “enable dictation”. On Android, go to Settings> Language & Input> Speech> Voice Input and choose either “Basic Google Recognition” (simple speech to text) or “Enhanced Google Recognition” (full hotword and interaction)
In addition to being embedded within the main mobile operating systems, voice recognition is also a core part of both Microsoft Office and Office 365. In addition to voice recognition software that comes bundled in either operating systems or other applications there are standalone software packages and apps specifically designed for this purpose.
It is fair to say that generally these are more sophisticated than the bundled options and they can do a lot more things that can automate even more of your processes. Conversely, they do need more learning to get the best out of them.
The dominant player in the standalone voice recognition market is Nuance. They offer a wide range of different systems from mobile apps through desktop applications and packages that work across all platforms.
The Nuance Professional desktop product actually won a “best in show“ award at the recent Empowering Technology Through Advice conference run jointly by FTRC and the Personal Finance Society, a video of their demonstration can be found here. This highlights how the additional functionality of the more sophisticated desktop products can, with the right training be incredibly powerful, not just as an alternative to typing but in automating many processes that are undertaken on a regular basis.
What you should know about voice recognition software
If you’re going to use voice recognition there are some important things to understand about how it works and the things that may go wrong. Computers don’t generally have the ability to separate words in speech. Humans have learnt that in general conversation one word can easily merge into another and our hearing adapts to this.
Twenty years ago if you were going to use the voice recognition product it was necessary to try to leave a clear gap between each individual word that you said so that you spoke in a rather staccato style. Modern systems no longer need this, but the truth is if you can take such an approach you will get even higher levels of recognition.
Those of us who used to spend lots of time dictating letters will remember techniques that made it easier for people transcribing them to follow, broadly the same things apply, try to speak clearly with precise diction and you will get better results.
Another really important thing to remember is that you don’t get spelling errors with voice recognition, but you do get “sounds like” errors. These can actually be very embarrassing and if you don’t identify them some of what you dictate may appear as gibberish. There are easy solutions to this that can also actually improve the quality of what you write overall.
All voice recognition software will have some form of text to speech capacity. This is where the system you are using reads the text back to you. It is very important to get the software to actually read back the transcribed text, not what you have dictated, although is also usually possible.
Reading back the actual text will not only identify “sounds like” errors but is also a really good way to listen to how your text will actually come across to the reader. In my experience this enables me to write more concisely with more effective punctuation and focus on the key messages I wish to get across.
One investment that is certainly worth making if you want to get the best out of a voice recognition system is a dedicated high-quality microphone. With increasingly sophisticated unified communications solutions, i.e. the ability to link your telephone, computer and mobile devices all to a single headset now being available, this can be really valuable.
For the last few years I have used various devices by Jabra that enable you to link different devices and move seamlessly between them. The best such systems also include really good noise cancelling technology. Plantronics and Sennheiser are another well known manufacturers with a range of options in this market. Which particular system will be best for you will depend on which desk-phone system you use, if you want single or dual ear headsets and if you want to be fully wireless. In my experience spending £300 or so on a really good headset is money well spent.
Voice recognition and data security
One issue users should consider is data security. If you are using the embedded voice recognition in mobile operating systems this means your data speech is being transmitted over the internet and transcribed on the servers at Apple and Google. It would potentially at least be theoretically possible for this to be interrupted and captured by bad actors, so if you’re going to use these tools it would certainly be prudent to have a VPN installed to secure all communications over mobile devices. That said using a VPN really should be standard practice on any mobile device used for business purposes. Mobile hotspots are notoriously insecure.
There is also the question of if you are happy that your dictation would be available to the company doing the transcription. Apple say they don’t make any use of customer data they obtain, however, obviously it is actually core to the way that Google work.
Data security issues certainly make a strong case for using the standalone products from companies like Nuance, who do argue that the confidentiality and security of their systems are one of the main selling points.
The biggest barrier to adopting voice recognition in your business may be that some people are self-conscious about dictating into a phone or headset in an open plan office. This makes no sense as we all speak into phones all day, but I have found some users find this a barrier.
The potential of voice recognition software
Overall voice recognition has enormous potential to deliver massive time savings to any business producing large amounts of regular emails and other documentation.
It could offer huge benefits to IFA firms; certainly those I know who use it regularly find it a great time saver for individuals and overall a way of greatly increasing efficiency across the firm.
What do you think about voice recognition software? Is it already helping you save time in your business — or are you going to try any of Ian’s tips? Let us know in the comments below.
Please note, the views and opinions expressed in this article are the views and opinions of the author and do not necessarily reflect those of Aviva.
