Slashdot Mirror
iPod Engineer Tony Fadell On the Unique Nature of Apple's Design Process
An anonymous reader writes "Often referred to as the godfather of the iPod, former Apple executive and current Nest CEO Tony Fadell played an instrumental part in Apple's resurgence. Recently, Fadell opined on what makes Apple's design process different from the rest of the pack. Fadell explained that a key and yet often overlooked, difference between Apple and other tech companies is that Apple ships 99% of the products that pass certain internal milestones. By way of contrast, during Fadell's tenure at Philips — where he was charged with overseeing the company's audio strategy — the iPod guru noted that Philips would axe 9 projects out of 10, even if a particular product was about to ship."
Sounds like that internal milestone is a special bar. How many projects reach that milestone? Is it more than 1 out of 10?
Excuse me for asking, but... How is "products that pass certain internal milestones" (aka Steve Job's early scrutiny) in any way related or comparable to "9 products out of 10, even if a product was about to ship"?
It would be interesting to know exactly what that internal milestone is. Obviously, the ability to axe projects is core to Apple's business, as evidenced by the tiny number of SKUs they offer at any point in time compared to most electronics companies. And there have been rumors that Jobs could be particularly brutal when it came to shutting down projects that he didn't think were worthy.
The difference must be that while all companies axe projects, Apple makes cuts earlier than other companies and only lets the few chosen projects make any progress in the lifecycle. Whereas other companies take a 'throw everything at the wall and see what sticks' mentality, and only cut projects later when they aren't good enough. Sometimes they cut too late (e.g. MS Kin).
Modern business have degenerated from organizations to make and sell products/services to support systems for management employees, CEOs, and the financial/banking sector to whom they all report to. They are in the business of business for the sake of business. Modern companies honestly see making and selling products as a nuisance that gets in the way of their real goal, which is making sure management gets paid and the stock price stays inflated.
OP's remark about 9/10 products being axed on a whim smells of the terrifying bureaucracy and labyrinthine organization that company must be. Microsoft has been rumored to be organized like a medieval kingdom with lords defending their territory with force, politics, and guile.
Personally, I expect to see an enron-like collapse of any number of large companies in the near future. The cause? Routing loops. Eventually every last function and service will be subcontracted and outsourced. Nobody will be able to tell who makes what, and where anything comes from. Eventually someone will realize that they've attempted to subcontract a product to themselves.. Many times over, the trail going dead after too many iterative loops.
The Odyssey Command Center (Odyssey 3) video game console was axed by Philips just as it was about to ship. It wasn't the strongest offering at the time but it offered backwards compatibility with Odyssey 2 games and was to be expandable with a modem and BASIC.
I was saving my dollars and ready to buy but it was axed shortly after they promoted the hell out of it at the CES and Knoxville World's Fair. Jerks.
Regards, Lex
What he's saying is that Apple has an actual functional interal milestone systems.
Other companies say they have a milestone system, but it's really bullshit lip service. What ships is up to the whim of whoever in charge, and failures are scraped under the rug because whoever in charge met the "milestone" and thus gets his bonus.
...Apple ships 99% of the products that pass certain internal milestones. By way of contrast ... Philips would axe 9 projects out of 10, even if a particular product was about to ship. ... "Nine times out of ten, or 99 times out of 100, they would kill the project, either at the beginning, the middle or right before the product was supposed to be shipped."
OK, I ready I read TFA - is this incomprehensible? Does it mean anything? Is there any useful data anywhere in this?
Three Squirrels
Sounds like that internal milestone is a special bar. How many projects reach that milestone? Is it more than 1 out of 10?
At Apple, the milestone was "Steve approved it." Everywhere else, it's decided by committee. That's why 9 out of 10 are yanked... just like anything else decided by committee.
#fuckbeta #iamslashdot #dicemustdie
The MS Kin was a special case.
It wasn't so much that it was a loser that should have been culled; rather, it was destroyed by poor decisions from Microsoft middle-management.
Basically, MS bought a successful company, Danger. Danger's "Sidekick" was a feature-phone with a well-chosen feature mix. Kin was to be the next Sidekick, and it should have been the same success the Sidekick was. The most interesting feature: it was supposed to have a special low-cost data plan. Instead of being a full smartphone, it was going to be a "social media" phone; SMS, Twitter, and Facebook wouldn't put too much load on the data network, so Verizon agreed to offer a special low-cost data plan.
Well, a Microsoft middle manager forced the guys working on the Kin to scrap the old Danger code base, and rewrite everything to Windows CE. After all, Microsoft didn't want to have to support two code bases, right?
But the delay caused by the rewrite was fatal. The special low-cost data plan evaporated (Verizon was pissed at the delays), and instead of being a low-cost phone with a low-cost data plan, it became a phone that cost about the same as other phones, and had a data plan exactly as expensive as other phones, but wasn't a smartphone so the built-in apps couldn't be added to. That last was really stupid: since the Kin guys were forced to rewrite to Windows CE, it should have been possible to put a Windows Phone app store on the device, and the Kin team wanted to do it. They were denied, again a stupid decision by MS management (and I guess internal MS politics).
Had the Kin shipped 18 months earlier, even 12 months earlier, with the less-expensive data plan? It should have been a big hit like the Sidekick. Had it shipped as a smartphone with an app store, it might have had some sort of a chance. But as a featurephone that cost like a smartphone, it was instantly doomed.
So yeah I guess MS should have culled it rather than endured the embarrassment around the Kin disaster. But better still they should have had less broken decision-making by their own middle management.
http://arstechnica.com/information-technology/2010/07/a-post-mortem-of-kins-tragic-demise/
http://www.engadget.com/2010/07/02/life-and-death-of-microsoft-kin-the-inside-story/
maybe not reworking but waiting for tech to catch up with the idea.... don't forget the iphone was born of the ipad.
i guess at the time, given current technology, Apple could't reach that milestone in a 10" formfactor, but they could in 4"
"I thought, 'My God we can build a phone out of this,"' Jobs said at The Wall Street Journal's "D: All Things Digital" conference in Rancho Palos Verdes.
Apple must have had the ipad idea as early as ~2000.. (phone launched in 2007, assuming 5 years in the oven... the idea came to him in 2002.. so the ipad must have been researched/prototyped at least a year ot two before that before they decided 2002/3/4 technology wasn't going to work for an ipad)... 8-10 years or so before the ipad actually made it to shelves.
Well, while you're thinking of the numbers, let's look at the product lineups. Philips has bajillions of products from light bulbs to shavers to stereos to all kinds of miscellany whatnot. So how many products were killed in development if this anecdote is anywhere near correct? Apple has the iPx mobile things, a handful of laptops and desktops, a server or two, and accessories for all the aforementioned. Do they have even 1,000 current products?
Whatever the exact number, the real point is that It sounds like everything at Apple is really tightly driven with a focus on only even bothering to start products that have a place in the lineup whereas Philips has a shotgun approach.
The big difference between Philips and Apple isn't whether projects are killed earlier or later.
The difference is how the projects come to be and reach these milestones.
Philips uses a "technology platform" system, or at least did during the time Tony was there. I don't know what they use now. That means someone in a technology division at the company develops a technology. Then they develop some platforms that use the technology. They then produce reference platforms or designs that use the technology. Then they take those reference designs around the company and try to find a product group in the company that wishes to ship a product like that.
The problem with this is that it is pushing a rope. You frequently will make up products that show off a technology but that few people would want to use let alone buy. This system was commonplace with companies at the time. You can still see this system if you look at something like dealextreme or meritline. You will see many companies (barely more than entrepreneurs in these cases) who make products simply because the technology lends itself to them, regardless of whether anyone would want to use it.
The big difference in how Apple did it, and still does it, is that Apple identifies a product people would want to use and doesn't currently exist or at least doesn't broadly exist in an easily usable form. Then Apple goes out and buys, develops or partners with a company to develop technologies that make that product work or work better. The company then evaluates the product before shipping it, deciding if the product is really something people would use. Rarely does the company have a change of heart about the basic product, but sometimes products get killed because the result doesn't really work in a way the customer would like it. For example, if a product doesn't work smoothly, it may be delayed until faster processors come along. The G5 MacBook Pro was fully developed and then killed because (among some other issues) the battery life was so short no one would find it useful.
And that's why Apple products usually ship, because they were designed to ship from day 0. Philips products started out being made simply because they could be, and so many of them died on the vine when it was realized no one wanted them or even if they just can't convince any product division they would like to ship that product.
Sources: I know people who worked at Philips. I have worked at Apple. And I've talked to these Philips people and Tony Fadell specifically about these particular differences between Philips and Apple.
Well, Dell ships more computers than Apple, as well. Samsung ships tons more phones, yes, but not many of them are their flagship ones. Every Samsung smartphone is called a Galaxy something, and they range from the completely free crap phones with crappy screens, to the S-III. Heck, Samsung just introduced their S-II something with a huge screen but... 800x480 screen.
So yes, Samsung better ship more phones, because they have probably over 50 smartphones in their entire product line, including ones that run Windows Phone, amongst others. Apple only had 3 models, 2 of which are laughable just to have a price point. Of course, Dell has a similar situation - they probably have hundreds of PCs, while Apple has what, 7 different ones?
These days, Apple's not about marketshare. Just the part of the market they want to make money on. (It helps that that part of the market is willing to spend money as well, because it's why iOS App Store is #2 in developer money (#1 is Blackberry, believe it or not), followed by Amazon App Store at #3 (about 50% of the Apple App Store). Distant last is Google Play - under 50% of what the Amazon app store brings.)
I agree that they launched a little bit too early. But the symptom of that was the unavailability of a native SDK at launch, and the associated nonsense about web apps being the platform. Also the original lack or cut'n'paste.
But not these...
For example, due to lack of processing/GPU power and a desire to make apps look slick they decided to go with a fixed resolution and mono-tasking. Now they are stuck with making every new screen a multiple of the original iPhone or iPad resolution, and suffering from black borders when they wanted to go widescreen. They can't easily introduce multitasking either, just a kind of bodge for a few select applications.
Neither of those are true.
Fixed resolution has nothing to do with limited CPU/GPU power. It's a positive design decision. That on a small screen apps need to be specifically designed for a particular screen, not be resizable.
And of course Apple could quite easily introduce traditional multitasking. It's intrinsic to the unix that underlies the OS. And all the levels above that were already created with multitasking ability, as they were adapted from OSX. And the CPU was certainly enough to support it. The iPhone CPU from the start was far more powerful than the original Mac CPUs that OSX ran on. The very easiest thing to do would be to introduce traditional multitasking.
They didn't for two reasons.
a) Battery life. You see it on Android very often that some crappily written app that's still running in the background takes hours off the battery life. That doesn't happen on iOS.
b) Simplicity of the UI for users. Phones are supposed to be simple devices, with app interactions typically being seconds rather than minutes or hours. Nor do Phones don't have overlapping windows, nor screen real estate for permanent docks/task bars - the indicators of multiple apps running on desktop OSs. So some other form of app switcher/manager is required for multitasking. The original concept was that this was too heavyweight for a phone.
The fixed screen size decision is a good one that has stood the test of time. iPhone apps ARE better for being specifically designed for the size of screen. And doubling is the perfect answer to higher resolution technology being available.
The longer screen size is fine, as in practice, the tricky dimension is the width. Most apps are list based, so having more of a list shown doesn't change the app design. Whereas changing the width would mean different text limits/layout of list items.
The initial design decision of no multitasking didn't last. But it's no bad thing to start with a very simple UI design, then add more complicated features later. And they did keep the battery conservation plan by only allowing system services at actually run in the background.
in the longer term they built a platform with many of the limitations that desktop operating systems suffered from in the 80s. Many never overcame those limits, and when they did it was often with a horror show like Windows 95.
Haven't a clue what you're talking about here. Presumably it's something about the lack of pre-emptive multitasking on early OSs. But the iPhone HAS pre-emptive multitasking. It just doesn't allow multiple apps to run arbitrary code at the same time. That's not the same thing.