A sad story with a happy ending? (and Comments)

10 July 06

My recent update on the history and prospects of Dragon NaturallySpeaking brought a response from John Shagoury.  He’s the President of Nuance Communication’s Productivity and Applications Division, which owns NaturallySpeaking.   He noted:

“Your analogy about crossing the chasm has some merit, but please don’t forget there have also been products that entered the market before the market was really ready, and the “hype” didn’t live up to expectation. In some of those cases, a re-invigorated “re-launch” at the right time actually achieved the astonishing success that the original “hype” was all about.”

John Shagoury, 26 June 06

John hopes that NaturallySpeaking is that kind of product.  I do, too.  I think the software is fabulous, and the product’s success would provide a happy ending to a sad and tumultuous history.

But my hope is tempered by experience and analysis, which indicates that the product will have a very difficult time achieving mass market acceptance.

For the last several years, the ICE Research Center has been developing and applying a simple analytical framework that aims to bring a higher degree of understanding to the causes of new product success or failure.  This framework looks at four attributes of a new product or service.  Two of these factors – higher benefits and lower price -- relate to purchase motivators.  The other two factors – ease of purchase and ease of use – relate to the amount of change required of potential customers to use the product.

Here’s what NaturallySpeaking looks like in this “four factor” framework, both for general consumers (the “mass market”) and for business customers that currently use dictation and transcription services – like doctor’s offices, for example.

Dragon NaturallySpeaking:

Moderate Motivators, High Barriers

The program asks an awful lot of its users.  They must learn how to use the new technology, and then spend time training the software so it can understand them.  This makes NaturallySpeaking a product that requires a pretty significant investment of time – say an hour or so -- before a customer can use it. 

These kinds of barriers are somewhat lower for businesses where NaturallySpeaking is substituting for another kind of dictation, as in the diagram on the right, since users are already accustomed to dictation.  Even in these businesses, however, the difference between voice recognition and transcription is significant.  Imagine, for example, asking an experienced doctor or lawyer to use a voice recognition program rather than his Dictaphone.

The big benefit for users of dictation and transcription services comes from the program’s economics.  NaturallySpeaking eliminates a measurable and costly task, transcription.  As a result, the attractiveness of the product as a transcription-substitute, in the diagram on the right, is much higher than it is in the larger consumer market, on the left, which has more varied and less quantified uses. 

Almost ten years ago, Harvard’s Clayton Christensen described a successful entry strategy for disruptive technologies – target the customers for whom the product performs a necessary job in a better fashion.  Initially, there might be very few of these customers, but if an enterprise can make a profitable business with a well-defined group of customers, it is in a good position to grow in later years.  As Christensen put it:

“Be patient for growth but impatient for profit.”

Clayton Christensen, Harvard Business School

 NaturallySpeaking could find a profitable market and business as a substitute for transcription.  This doesn’t put the program on every desktop, but allows it to focus on a group of users who get tangible and measurable value from the software. 

For the consumer market, however, NaturallySpeaking’s profile looks less encouraging.  True, advances in technology, both software and hardware, have made the program’s operation much better than it has been in the past.  But at the same time, there’s been a proliferation of the ways in which people get information into their computers. 

Much less of this information is entered as sentences and paragraphs.  Much more is entered in a text-based language that relies on abbreviations and artful mis-spellings, both because of increased demands for mobility and for speed.  Voice recognition technology can’t help with a statement like this one:

wut r u doing 2nite

Instant Message, courtesy of The Christian Science Monitor

NaturallySpeaking missed its moment for mass market success.  In the absence of viable speech recognition technology in 2000, many potential customers adapted by changing the way they used text to communicate.  Now that much better speech recognition is available, the need for it has declined.

I responded to John’s email, which I quoted at the beginning of this update, by telling him that I hoped my prediction was wrong.  Personally, I'm a big fan of NaturallySpeaking.  In fact, I used it to write an early draft of this update.

But, based on the four factor analysis above, the prospects of the software recapturing the imagination of a large market seem to me to be remote. 

I’ll check back with John in a year or so to see what is happening with NaturallySpeaking.  If it attains the breakthrough success it deserves, it would be a fairy-tale ending to what has, thus far, been a sad story. 

More Information:

1. A description of the four factor model for new product success is here.
2. Clay Christensen’s web page is here.
3. Last week’s update on the history of NaturallySpeaking is here.
4. Here’s the homepage of Nuance Communications, Inc.