Endeavour Partners

I saw this at the gas station yesterday; I suppose they have one at every gas station:

Air is free, but water isn't

It struck me that at one time or another someone may have thought they would make money from air.  Or clearly, that they were monetizing this device in some other way.  For instance, by attracting customers from the station across the street that didn’t have free air.  But now all the gas stations have this service, so no one makes any money from it even in an indirect way.

It’s also interesting that air is free, but water is expensive at the gas station.  In fact, water is more expensive by the gallon than gasoline!

We’ve been thinking a lot lately about why some things are free and some things are expensive – and, in particular, why so many more things are becoming free.  Technological progress, asymmetric competition, and open innovation are relentlessly driving costs down and capabilities up.  A prior post pointed out the innovation trajectory of processors, storage costs, and bandwidth.

Asymmetric competition is when competing players have very different motivations and are happy to destroy each others economics in order to chase these goals.  For instance, Google has been going along merrily, annihilating the value propositions of email, watching videos, smartphone operating systems, office productivity applications, and mobile navigation all in the name of capturing more eyeballs and clicks for advertising.  These are all areas where people once spent billions of dollars.

A recent video I saw of an NEC product under development brought this all home to me, and it is worth looking at more closely.  NEC appears particularly blind to the power of open innovation – leveraging the achievements of others and the vast army of technologists and developers moving forward at breakneck speed in a wide-variety of areas.  Instead of building on the best of what is available, they appear to be developing something proprietary and from scratch.

My friend Josh at 3Play Media showed me the video of  this NEC product, a cool new “universal translation” system.  Well, at first it seemed like a cool idea.  Two users with glasses.  One user speaks in, say, Japanese, and then the other can see the words in English projected on the lens of the glasses.  Wouldn’t it be great if you could go anywhere and have the world translated for you like this?

The system has two glaring problems, however:

1. The price – NEC is targeting a price of around $83,000, making it barely competitive with hiring a full-time human translator
2. The clunky hardware – the current lab version of this is so geeky and clunky as to be completely ridiculous and impractical at any price

Of course, NEC plans to bring the price down with volume.  And I am sure that they will be investing to make the glasses nicer and more like “regular” glasses.  Unfortunately for NEC, they seem to be investing a lot of money in things that might not matter that much (like wearable computing) and competing against an avalanche of free stuff that likely will.

The universal translator has five key elements:

Five elements to make a universal translator work

1. First, we have the device itself.  The concept of a wearable, glasses-based system is intriguing and sexy.  The reality, however, is horribly expensive, clunky and a long, long way from ready for mass market appeal.  Also, why have a specialized device for translation?  Wouldn’t the use case be nearly identical if you presented the translation on an iPhone or Android device or even better if you text-to-voiced it to a blue tooth earpiece?  These technologies are here today and effectively free for this service.
2. Then we have some part of the service working in the cloud.  Sounds good, but don’t expect to charge extra for this.
3. Then we have voice recognition.  Similar to the device, don’t we have armies of developers working on this problem in other fields?  Won’t voice recognition just improve on its own and not require NEC to spend much time or money on this element?  Eventually, there will be a voice recognition piece that can be bought off the shelf and work well enough.  It won’t cost $83,000 and won’t contribute to NEC’s value creation.  Windows 7 has apparently made some strides in this area.
4. The translation itself.  This has to be the secret sauce, right?  But once I have it in text I can already go to Google and translate it for free.  Any language into any other language.  No, it isn’t perfect, but is it good enough to avoid paying much?  Or do I want the $83,000 version?  Probably good enough.
5. Presentation of results of the translation.  This is a solved problem, very easy in either text format or voice.

So, what is NEC basing it’s pricing strategy on?  And when are they planning to come to market?  Based on this quick and dirty analysis, we should have a free app on iPhone within a year or so (if there isn’t one already!).

I guess this means translation is more like air than water.