In the comments on my post about the non-death of the tablet, Christopher Coulter writes:

But the only focus is large enterprise workers who are in "meeting to meeting" mode? The so called, (but badly named) Corridor Warriors? But why narrow it to only "large enterprise info-workers", as Pen Computing is meant for everyone. The pen is a natural interface. Education, Small Business, Artists, Architects, Engineers, Form-Fillers, POS Units, Science, Peace Officer/Military, Health/Care and other strict Verticals, End Users, Government/Groove types, Info Workers/Corporates, Kids (yes, that's a market. You should SEE my 5 year old nephew. They take to Tablets [Pen Interfaces] like everything), Gadgetheads, Bloggers and other (webby algae-lifeform) Geeks, Writers (Scriptwriters), Franklin Covery types, bascially everyone. I totally agree with your non-techie thrust however, just curious as to why usability testing was limited to only a certain demographic. Where's all the rest of the Amy Tan's and Rob Lowe's?

And here's the thing, tech comes into "large enterprises" mainly via the back door, grassroots if you will, non-techies using it for own purposes that have direct impact upon the market of said Enterprise, on that I speak from personal experience. But to microscopically hair-split markets into easy-definable "target user" and "usability tested" objectives doesn't always mirror the real world; thats MBA 'planned-economy' rose-colored glasses, the world is far more complex and "laissez faire". To ignore End Users and people outside Large Enterprises and to ignore the Buzz and the Media/Analyst kickbacks, kills markets before they generate. To place all chips all on the 'well, see, it wasn't really meant for you, you all are just Early Adopters' is to me, a cop-out. :)

This is a great comment in that it's exactly targeted towards “theory“ of product creation that I personally love.  So first some background on building usable, needed and useful interfaces. 

Alan Cooper has a theory that he expounds in his books particularly the “Inmates are Running the Asylum“ where he talks in theory about how interfaces are over-developed in that they are being built often times to please too many users. We have to have the Blue widget over there for the democrats, and the Red one for the republicans, and oh, of course can't forget the green party.  And while we're at it we should have widget for tall people and short ones.  Oh and now that we've finished the design, we forgot about the one for the libertarians, etc.  In the analysis, Cooper argues that you should design for one persona (or a very small set of personas) almost exclusively since if you can make that one persona extremely happy that one persona will have an interface that they love and will gladly use it.  Thus the idea is that the one persona you pick represents the target market you're after.  The theory basically goes on to say that if you got that one persona 100% satisfied you're probably going to get other personas satisfied as well, but clearly not at the 100% level.  [Please don't criticize me on the theory, it's not mine personally and I probably did a poor job of paraphrasing it down to one paragraph]

This is counter to Microsoft's and many other software development strategies which basically tries to build something for 80% of the users at a much lower satisfaction rate.  Ever since I've been at Microsoft, I've heard program managers and others reference the 80/20 rule as part of their decision making process.  To me the issue here is that I agree in principle with Cooper in that trying to satisfy 80% of the people 80% of the time in 80% of the situations really results in a relatively poorly designed interface that's designed to do everything for everyone and just does it poorly.

Apple at some level does it right, but only for a very different reason.  Steve Jobs has an eye for design and oversees at a very micro-level the design process of the machines and the interfaces.  In talking to some former Apple designers, including the design lead for OS X - the Apple method is that everything is designed to please Steve.  Thus in theory they're using a persona “Steve Jobs“ as the representative of user market as they're trying to get to 100% satisfaction rating for this one person.

Okay so I've taken a bit of a diversion here, but back to the main thought.  When I started working on the Tablet PC project it was clear to me that long-term success meant having the right basics for the eventual target market of our users.  This is the key in my post-rationalized thinking.  Remember how I did analysis and worked on pen computing on the past?  One of the important insights here which I probably didn't really spell out in the previous post was that it was that these markets had all the buzz, had plenty of early adopters and evangelists - but what they didn't have was a well target core of functionality that would let them into the meat of curve where early adoption leads into the mass market of users.  To me this is why many of these products failed, not because they we're great invention, but rather the complexity and fundamental raison de etre' was not apparent.

So in Tablet, I threw our focus in making sure that those late adopters would have the fundamentals and that it would be easy enough, useful enough and needed functionality when they get around to having one of these devices.  This isn't to say that the tablet team ignored the early adopters or that they are ignoring them now, but rather if we could keep the team focused on this one set of users now (the corridor warrior -- I'll take a little bit of the blame for that one too) we'd have the foundation in place for later.

The other piece of the puzzle to realize is that in studying early adopters/adoption most of these users will put up with horrendous interfaces and limited functionality because they can see the promise of something cool coming down the road.  So as a result we didn't spend a lot of time worrying about them (from a development perspective) since often they take care of themselves.  From a marketing perspective, I'll probably agree that we didn't necessarily do enough.

Now as to all of the market segments from verticals to education to the creative types.  We simply didn't have the time or the resources to study in-depth what these people really need in a pen computing interface.  I'll write another post in a couple of days where I'll talk about the depths we went to understand the so called corridor warrior - the efforts here were huge, but trying to replicate that level of understanding for these other markets wasn't necessarily in the initial set of cards.  But also at the same time we wanted to leave the door open as a platform to let other application vendors to come in here and build products in areas where they have the expertise.  People like Arin and others on the Tablet team are actively working with ISVs to help them build on the platform so that these applications can meet the needs of the different segments.

Thus the evolving approach was to get the basics right for the mass market (horizontals), hopefully give enough for the early adopters to see promise and build enough of a platform to enable others to expand it (verticals).  How's it working... I guess we'll see as time goes by.