2011.32 | Speech Recognition @ Retail

Speech Recognition

I really enjoyed reading over this recently released infographic on speech recognition.  Things have come a long way since a keyboard challenged office mate of mine bought a headset with attached microphone and commenced annoying my area of the office with his verbal discussions with his word processor over a decade ago.

Google’s iPhone app has had speech recognition capability for ages now, but I don’t really use it when I can just type something in – most searches are short enough that speaking it isn’t worth it.

I also have used Dragon Dictation for iPhone and iPad, but I’ve found I need to change my thought process for text entry.  It’s useful for a slightly longer text message if one is alone, but I don’t find that I speak out loud in the same way that I write, so the dictation app is not as useful for my use of  jobs like word processing – or blogging.

While I don’t find it useful for writing or searching, speech recognition is extremely useful in the car, when you cannot access your phone while driving – it is illegal in my area, so hands free operation is a must.  With solutions like Blue Ant’s S4, connected to Vlingo it is even possible to hear and send text messages and more via speech to text and text to speech.

In fact, from a retailer and consumer facing organization’s perspective, the biggest opportunities for voice recognition could be for using in an automobile scenario.  As per my last post, 2011.31 being able to order while driving by talking to the car is a twist on the mobile channel that provides an increased level of convenience for time constrained consumers.

Another interesting angle is the potential of providing speech recognition solutions in a drive thru environment for quick service restaurants.  While having customers use a touch screen would slow the order taking capability, if speech recognition improved enough, it could provide a potential throughput improvement for a drive thru.  Instead of relaying items to an attendant who lists them on the screen at car side, speech recognition could do the same thing more quickly and ideally more accurately – showing the customer items on the screen as they are spoken.  Sites with multiple lanes could be monitored by one person with improved throughput.

To take it further, this system could even be extended away from the outside of the restaurant.  Clients could dial a number for their local Quick Service Restaurant from their car and put together their order via speech recognition, and be given a unique and memorable order number –  their mobile number perhaps.  When they arrive at the drive thru, they can identify their order with the number.  The order can be pulled from the POS system as a suspended order, and the customer proceeds to the payment and pickup windows.  This would allow the customer to speed their ordering, and the restaurant to increase throughput at the busiest part of many locations.

As always, there would be a great deal of tweaking to be done, particularly around operations.  For example, if items are out of inventory, the system needs to know so that customers cannot order items that are out of stock, and ideally alternatives are provided to the customers.  In order to prepare this for prime time, some serious stress testing would need to be completed.

Speech recognition is far from optimal at present but has improved incredibly.  We have all dealt with IVRs that frustrate us, for example while trying to get help on a hotel reservation or change our phone service, but as it improves some intriguing possiblities reveal themselves.

Image Created by: Medical Transcription

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: