« Anti-TV follow-up | Main | This blog is under Creative Commons »

The Concept Carification effect


In the cover story in this week's Time magazine, Steve Jobs talks about "How Apple Does It." One of my favorite parts was this:

"Here's what you see at a lot of companies; you know how you see a show car and it's really cool, and then four years later you see the production car, and it sucks? And you go, What happened? They had it! They had it in the palm of their hands! They grabbed defeat from the jaws of victory!
"What happened was, the designers came up with this really great idea. Then they take it to the engineers, and the engineers go, 'Nah, we can't do that. That's impossible,' And so it gets a lot worse. Then they take it to the manufacturing people, and they go, 'We can't build that!' And it gets a lot worse."

Those car pictures show the before and after of the Chrysler "Turboflite" concept car. It's rather obvious that the "after" car, from 1965, looks nothing like the 1961 concept car. What happened?

And the same thing happens everywhere. There is a major computer book publisher (not O'Reilly, as will be obvious), where this guy (author/editor) had a wonderful concept for a new kind of computer book. Not like Head First, but every bit as unique and engaging. He had a vision, a manifesto even (I don't want to link to it or mention his name because I don't want to get anyone in trouble here). Authors were excited, people were on board, and the first book began production. You know how the story turns out, since it's the same story that plays out all too often... the very thing Jobs described. The production people started saying, "Oh, we can't do THAT..." and the resistance piled up until the book was released looking virtually like every other book, save a few fonts and a very weak theme. The guy with the original vision was disheartened. One of the other original champions of the project left the company, partly as a result of watching this concept have the life and sharp edges sucked out of it.

One of the things we loved about O'Reilly is that they said, "Yes, do it ALL." The Head First format is virtually identical to the concept Bert and I built in the original proposal. No edges were smoothed. Nobody said "we can't do that."

Obviously there are a zillion reasons why wild-ass concepts can't (and shouldn't) find their way into final production, but how many of those reasons are truly valid? When people say, "We can't afford to do it that way..." we should always ask, "Can't... or Don't Want To?" followed by, "Can we afford not to?"

If being remarkable is one of the only ways we can hope to compete in a world where everything has a ton of competition...

Of course, the article goes on to talk about how Bill Gates has "kicked the bits" out of Apple, proving that there IS another way, and that this way can be more successful. Which leads to my REAL favorite part of the article:

"Jobs doesn't care just about winning. He's willing to lose... He's just not willing to be lame, and that may, increasingly, be the winning approach."

We have to keep fighting the Concept Carification effect, to keep at least some of our ideas alive, sharp edges intact. This is not an easy battle, since it involves separating the crap ideas from the brilliant concepts, with NO evidence. After all, most revolutionary concepts do NOT come directly from what users ask for. That's where we need to have faith. Yes, there are a ton of crap things out there that should've stayed in the concept stage, but if that's the price to pay for a world in which not everything is morphed into a nice safe incremental release, it's so worth it.

So have faith. When you're really really on to something magical, you can guarantee there will be devil's advocates, naysayers, and viscious critics every step of the way. Yes, sometimes those critics will be right, but if we aren't brave enough to fight through it when nobody knows for certain, then everything good will be stuck in the concept stage, and we'll be left with... all of the boring, undifferentiated, or lame products we have now.

Posted by Kathy on October 21, 2005 | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference The Concept Carification effect:

» Killer Ideas vs Idea Killers from achievable ends
As mentioned elsewhere, Kathy Sierra is one of my favorite writers. Her Creating Passionate Users is a must read blog. The Concept Clarification Effect talks about how great ideas are too often destroyed in the process from concept to [Read More]

Tracked on Oct 22, 2005 10:43:57 AM

» Killer Ideas vs Idea Killers from achievable ends
As mentioned elsewhere, Kathy Sierra is one of my favorite writers. Her Creating Passionate Users is a must read blog. The Concept Clarification Effect talks about how great ideas are too often destroyed in the process from concept to [Read More]

Tracked on Oct 22, 2005 11:24:36 PM

» Faith in the Concepts from nuff blogged
"Creating Passionate Users" links to a Time article about how Steve Jobs and Apple have maintained their success in a Microsoft dominated market by staying true to their concepts--even if they seem foolish and unreasonable. In it, they say that compan... [Read More]

Tracked on Oct 23, 2005 6:05:35 PM

» Killing Creativity from Library Autonomous Zone
Creating Passionate Users: The Concept Carification effect by Kathy Sierra on October 21, 2005. Kathy Sierra is one of the strongest advocates for users of anyone I know. Her "Head First" books from O'Reilly are best sellers because she knows how to wr... [Read More]

Tracked on Oct 24, 2005 10:32:48 AM


OK, it's hold-up-the-mirror time.

You guys sharp-edged, your-vision-undiluted great books, but are you making them on concept carified subject matter? You're making the best books on Java and EJB's and etc., but to more and more people (IMO) it seems that Java and EJB's and J2EE are starting to exemplify the very principles you decry. Java is having more of a "designed by committee" feel. Does anyone feel like they kicked ass after making a EJB?

I'm finding it very interesting what Bruce Tate has been doing, someone whose book career (if not his whole programming career) has been based on Java now seeming to turn his back on it.

Don't you guys feel like you're missing out on something sticking with Java? Isn't there something to the more dynamic languages, and couldn't they use your approach to writing?

If you have a quick look at the URL I entered you'll see where I think a good starting point would be, but it may be too much to ask.

Posted by: Richard Cook | Oct 21, 2005 1:27:45 PM

Richard: ouch. But you're right. Notice we haven't come out with any new Java books this year? (just updates for Java 5) You couldn't be more right about EJB, either. Personally, creating EJBs (in their current incarnation especially) is usually an "I suck" experience. I still love Java, and to be perfectly honest -- in the current ultra-depressed tech book market, Java book sales outnumber the sales of *almost* all other languages combined. I can't pay the rent without book sales, so there's that... but Bert and I are both in a more risk-taking mood right now, so we're taking on non-Java topics, and a new series/format (including the very different "passionate users" book).

The very next Head First book, out in about 6 weeks, is from Eric and Beth on HTML/CSS/XHTML)

"Don't you guys feel like you're missing out on something sticking with Java? Isn't there something to the more dynamic languages, and couldn't they use your approach to writing?"

Alright, alright you sold me! ; ) Stay tuned...

Posted by: Kathy Sierra | Oct 21, 2005 1:40:51 PM

Sounds like a bit of a baby/bath-water problem here to me. Sure, entity beans are a mess, but Java is still a pretty darn cool language IYAM. Now (((Lisp))), on the other hand...??? *kidding*

Honestly though, I don't see that the Java example really applies to this concept car thing. Java may not be the new shiny thing (like Ruby, e.g.), but it's not as though it's being stripped down to something simple and boring (quite the opposite).

(BTW, Richard, I did a quick search and didn't have much luck finding Tate's blog (seemed down?)...could you provide a link supporting your comment about him? I'm curious to see where he's headed. --thanks)

Posted by: Dave Wood | Oct 21, 2005 2:50:23 PM

Dave: aren't you supposed to be finishing *your* Head First book? What are you doing commenting here? I'm not willing to say that Lisp is the concept car of languages... (but hey, I actually thought I'd be a Prolog programmer at one time, so what do I know), and the example is a *little* stretch... but here's what Richard's comment got me thinking about that I hadn't been considering in all this talk of initial innovation--What happens when a product does release in close to concept-car state--with all its edges--but then over time has its sharp edges worn down, or too many weird edges that weren't in the original concept get retrofitted in Frankenstein fashion, etc. What happens when the original concept car that *was* released starts losing more and more of its concept-carness? [I am so making up stupid words here, I know]
And Dave, you're the first one of us to write a Head First software development book that *does* add in at least *some* code from other languages besides Java like... Ruby : )
[FYI for readers here -- Dave's working on the Head First Objects book]

Posted by: Kathy Sierra | Oct 21, 2005 3:12:22 PM

Next in the lineup:
Head First Rails?
Head First Agile Development?
Head First Io?

Dave: this very long thread at the server side would be a good place for Tate's ideas about where Java is heading:

Posted by: Rob | Oct 21, 2005 3:20:00 PM

I think when Java first came out it was more in line with the Concept Car made real. It made a lot of tough choices. Leaving things out resulted in a beautiful, elegant design with some real innovations (synchronize!)

Scripting languages seem to have have a built-in early-adopter bias that makes them take flight quickly. They also seem to be a good choice for small, short lived projects. But, their dynamism is often a problem over the long term. You get a lot done early, but all the while you're building "technical debt" that eventually makes it impossible to make changes, because nobody left around understands how the code works, and the code doesn't offer a lot of clues, because there's no typing.

Perhaps more modern scripting languages like Python and Ruby will break this pattern.

We've never taken to EJB, so I can't comment on that, though, we always stayed away because it seemed like NO FUN.

Believe me, since it came out I've had many, many I RULE moments using Java.

Posted by: Charlie Evett | Oct 21, 2005 6:00:08 PM

I have to say, going to that URL with the Time article on Apple, and only being able to read the first paragraph was definitely NOT an "I Rule" experience. :-( Damn. I really wanted to read that.

Posted by: Beth Freeman | Oct 21, 2005 6:14:38 PM

Since other people brought up the idea that the head first books are great books on old topics and they suggested a few new topics, I had to chime in with a few suggestions of my own. How about:

* Head First Spring
* Head First Advanced Design Patterns
* Head First ORM

Just to name a few.

Posted by: Chris Johnston | Oct 21, 2005 9:51:14 PM

OK ... at the risk of starting a flamewar, I'd say "Don't do 'Head First Rails' or 'Head First Agile'" for a couple of reasons. First and foremost, there's room in this world for both Head First and Pragmatic Programmers, and I think the folks over there (Pragmatic Programmers) are doing what they do better than anyone else could.

Second, I don't think I'd buy Head First Ruby or Head First Rails or Head First Agile because I can learn from anything that's technically accurate, especially if I can do it hands-on, which open source allows you to do. Heck, you can even join in the fun and contribute.

You know what I'd do if I were in your shoes? I'd do "Head First .Net". When I go into Powell's Tech Bookstore, I see acres of books on Java and acres of books on .Net. Way more than Perl, PHP, Python and Ruby and LISP *combined*. Programmers love Perl, PHP, Python and Ruby, but the guys who pay their salaries love Java and .Net.

The Concept Cars of languages -- well, LISP 1 (the original McCarthy language) was certainly a concept car, with LISP 1.5 being the real thing. Anyway, I really think there were only about half a dozen truly unique programming languages -- macro assembler, FORTRAN, LISP, APL, FORTH and SmallTalk. Those were the concept cars.

And look what we use today: FORTRAN and LISP are still going strong, APL has an open-source spinoff called A-Plus which is a "cult" language, FORTH has one strong vendor (Forth, Inc.) and another cult. I'm not sure anyone codes in macro assembler any more, but people still write macro assemblers. SmallTalk? It has another cult spinoff called Squeak.

Posted by: M. Edward (Ed) Borasky | Oct 21, 2005 10:53:24 PM

That is so dead-on. I've just spent three and a half years working for the very company you just described. I managed to push innovation through layers of apathy and mediocity, but much of the coolness of the original designs was lost in the process. Nothing makes me angrier than this in the product development world. Bad leaders create bad teams. Bad teams create bad products.

It all leads back to the person at the helm.

Great post, as usual.

Posted by: Olivier Blanchard | Oct 21, 2005 11:16:31 PM

I didn't intend this to be a "which head first books should we do?" topic, but... I want to address Ed's comments:

I agree with you about *almost* everything. First, as far as I know there aren't any plans for a Head First Ruby and/or Rails book, but I'm actually *not* doing Head First books any longer (except helping with a couple that are in progress).

Eric and Beth have taken over the Head First series; Bert and I are working on a *new* series/format, which the passionate users book is *partly* using...

Second, I completely agree with you that the Pragmatic guys do what they do better than anyone else. The question would be if there's something in the Ruby and/or Rails world that targets a *different* audience (or a different audience need) than their books, and I'm thinking about that -- but not for Head First. So if there's a *complimentary* book on Ruby or Rails, that doesn't attempt to duplicate what the Pragmatic guys have already done superbly well, then I'd like to hear what you guys might want! Because I have to say, each day I get a little bit more excited about Ruby ; ) A little like that 1996 Java feeling again...

"Second, I don't think I'd buy Head First Ruby or Head First Rails or Head First Agile because I can learn from anything that's technically accurate,"

This is the sentence that doesn't make sense to me in this context -- I completely believe you -- but that means you'd never be a candidate for *any* of our books, regardless of the topic, so how does this apply?

We don't DO books for people like you : ) precisely *because* you don't need it, and we're trying to help people who *do*.

The brain-friendly techniques are for the people who--like me--have a much harder time with that. And usually, these are the folks who come to a technology a *after* the first wave of early adopters. Tim O'Reilly has found a consistent pattern where books for advanced people are more heavily favored in the beginning of a technology, and as it matures, the beginning books take over.

And though I appreciate any suggestions, I'd rather shoot myself than do a .NET book ; ) I MUST feel some passion for what I'm helping people to learn, and it's just not there for me in .NET.

But all of us here -- Eric and Beth for Head First, Bert and I for other things -- are definitely open to *any* feedback, input, suggestions. But while there are a TON of great topic ideas that we'd love to do, we also have to face the economic realities that the tech book market has plummeted (and I mean way, way, way down), and even with all of our books as top bestsellers in the largest categories, we're still struggling. We believe it isn't that hard to make a bestseller, but bestseller no longer means much money. We don't *do* this for the money--we do it because it's what WE are passionate about, but we need to make enough of an income from it to be able to continue doing it.

However, printed books are definitely not the future, although they will always have an important place. And I don't believe eBooks are the future either, because until screen resolution and portability is as good as paper -- they basically suck for all but short things. I cannot comprehend someone trying to read Head First on a screen! (Reference books, sure, but not start-on-page-one-and-keep-going books).

No, there are other kinds of learning experiences that could be very cool... and I'm particularly interested in *those* ; ) Let's just say that we're thinking about a learning "concept car", and intend to build it without losing the sharp edges (which is still no guarantee that it won't suck), but we aren't giving up books either. We LOVE doing the books, and we still have to stay alive, so once we finish our current books, we're going to split our time -- half on books and half on writing software again for the concept car.

Posted by: Kathy Sierra | Oct 21, 2005 11:56:59 PM

Its impossible to describe how cool your site is! Its just one damn fantastic post after another.

Im a teacher... but find precious little inspiration in the education field. But youve given me an easy fix... I simply change "user" to "student"... and find my educational worldview transformed.

Keep writing!

Posted by: AJ | Oct 22, 2005 2:20:36 AM

I'm not a programmer. Ruby on Rails sounds like a nice drink in the dining car. Ajax is a crummy suburb of Toronto and Lisp... well I won't go there.

You are one of the few blog writers who consistently provoke, challenge, entertain and educate. I don't know where you find the time - but I'm very glad you do. I look forward to the new book.

Posted by: Bill Kinnon | Oct 22, 2005 10:57:52 AM

I *really* liked this blog posting. It fucking rocks to just come out and say you aren't okay with mediocrity; turning visions into reality involves figuring out how to do the impossible. Life is short, so why do anything else?

Posted by: Brad Neuberg | Oct 22, 2005 5:13:55 PM

I'm reading (or trying to read) Scott Ambler's "Agile Database Techniques" book, and if there's any subject matter that needs the "head-first" treatment to bring those learnings down-to-earth and grokable, that is it. I would write it myself I if I had more expertise in that area.

Posted by: keith ray | Oct 22, 2005 10:39:12 PM

The full article about Apple is available here fo' free.


Posted by: Nivi | Oct 24, 2005 12:17:13 AM

A Head-First book on HTML/CSS/XHTML? I must be on Santa's 'nice' list for this year!

Also, is Java still really outselling? Is there a downward trend? When I go into the local Borders' and Barnes & Noble's I'm seeing the Java section shrinking. I'm not sure what's growing, though. I don't think it's one topic, but spread out among things like LAMP subjects, some Python, Ruby, digital photography.

Posted by: Richard Cook | Oct 24, 2005 8:07:42 AM

One last thing...
"And though I appreciate any suggestions, I'd rather shoot myself than do a .NET book ; ) I MUST feel some passion for what I'm helping people to learn, and it's just not there for me in .NET."

A good article topic would be "how to be/teach passionately about a topic you're not passionate about." Some of use may be at places we can't get out of, or (like me) we like what we do, but occaisionally have to work or teach something we don't really care about.

Posted by: Richard Cook | Oct 25, 2005 8:15:24 AM

I like the spirit of this post, but Ihad to laugh when I saw the concept car graphic. Hard to imagine who was going to buy that concept car, let alone drive it (does the top blow off at high speeds?)

Alan Cooper's bit about implementation models vs. mental models is pretty useful here. You want to strike a balance as close to the ideal concept/mental model as possible while still making something that is implementable. These days I'm finding "implementable" refers as much to *business* constraints as it does to technology constraints.

So, yeah, don't lose the cool concept, but don't get married to a lemon either...

Posted by: gretchen | Oct 25, 2005 5:36:27 PM

The comments to this entry are closed.