Saturday, August 1, 2009

The Steve Jobs method

Image representing Steve Jobs as depicted in C...Image via CrunchBase

It's been a long time since I did a post that was primarily a link to another blog with commentary, but I came across something today that I really want to share. One of the most common questions I get about the lean startup methodology is, "but what about Steve Jobs?" When I try to unpack what people mean by the question, here's my best take on what they are asking: "Look, Steve Jobs doesn't go out and ask customers what they want. He doesn't put out crappy, buggy products and then ask for feedback. And he doesn't shy away from big-bang launch events. He tells customers what they want, and he gets it right. So how do you reconcile his success with the lean startup, which seems to suggest the opposite?"

I rarely give a satisfactory answer to this question, because I don't know Steve, nor have I worked at Apple or Pixar. So I can't speak for what happens on the inside. Luckily, neither can most of the questioners who pose that conundrum to me. We all seem to have a mythical sense of how Jobs works, based mostly on speculation and our very human desire to believe in heroes.

My normal answer is that I don't really think that's how Apple products are built. Plus, the premise of the question misunderstands the lean startup, too. Like any good pundit, that lets me pivot back to talking about something I do know about.

So imagine my delight when I saw this blog post with excerpts of a Steve Jobs interview. Here's the key quote:
Steve Jobs on why Apple doesn’t do market research - Bokardo
It’s not about pop culture, and it’s not about fooling people, and it’s not about convincing people that they want something they don’t. We figure out what we want. And I think we’re pretty good at having the right discipline to think through whether a lot of other people are going to want it, too. That’s what we get paid to do.
The key phrase for me is "having the right discipline to think through whether a lot of other people are going to want it." That's what so many techniques that I advocate are all about: customer validation, minimum viable product, vision pivots, and even throwing away working code. Getting customer feedback is emphatically not about abandoning your vision or abdicating responsibility for innovating. Instead, it's about testing visionary ideas against reality, to discover what really works. Put another way, feedback's not about you - it's about them. When a customer tells you how they feel about your ideas, that doesn't tell you anything about your ideas. It tells you something about what that customer thinks and feels. Figuring out whether and how to incorporate that new information into your vision is your job. As Steve says, "That’s what we get paid to do."

Now, I can't speak to what process Steve Jobs uses to get his team to do this market assessment. Maybe they do it at the whiteboard. Maybe they just have great gut instincts. Or maybe there is the occasional potential customer or early prototype involved. But I'm willing to make some guesses. Here's how I make sense of their success. From here on out, this is strictly my imagination talking. To be clear, I don't know if Apple really works this way.

First, note the important use of work-in-progress constraints (kanban). As Steve says in the source interview:
Apple is a $30 billion company, yet we've got less than 30 major products. I don't know if that's ever been done before. Certainly the great consumer electronics companies of the past had thousands of products. We tend to focus much more. People think focus means saying yes to the thing you've got to focus on. But that's not what it means at all. It means saying no to the hundred other good ideas that there are. You have to pick carefully.
Having so few products means Apple can dedicate enormous resources to each project once it gets the green light. But it also means they have to be very careful kill projects if they are not trending towards something great. Which comes to the second major principle: halt work that leads to more waste, even if it means abandoning sunk costs. This is a version of the andon cord technique from lean manufacturing. Steve describes it like this:
At Pixar when we were making Toy Story, there came a time when we were forced to admit that the story wasn't great. It just wasn't great. We stopped production for five months.... We paid them all to twiddle their thumbs while the team perfected the story into what became Toy Story. And if they hadn't had the courage to stop, there would have never been a Toy Story the way it is, and there probably would have never been a Pixar. "We called that the 'story crisis,' and we never expected to have another one. But you know what? There's been one on every film.
These two principles combine to free up tremendous resources for raw R&D and innovation, because so few people are stuck working on "death march" internal projects or maintaining low-success released products. What do all those other people do? For one, capacity development. Apple has a track record of creating lots of interesting enabling infrastructure, like Quicktime and Bonjour, which sometimes become key to their products and othertimes not. My guess is they have lots of people constantly working on interesting new tools for their designers to play with. Those new capabilities must translate into a constant stream of prototypes - most of which turn out to be utterly bad ideas.

The real question is: how do they evaluate a prototype to know if it's a good idea? If they were an unknown web 2.0 startup, they could release the app and see how customers respond. But that would be a disaster for Apple, so whatever testing they do has to happen in secret and behind closed doors. (For startups that are tempted to mimic this behavior, I suggest reading the great account of the early Apple in Founders at Work.) Regardless, they must cull a lot of bad ideas for every one that we hear about. Holding his team to a high standard for what constitutes a great idea is what I imagine to be the most value-creating part of his job.

Most executives, especially in startups, don't have the courage to hold their teams to a high standard for new products or features. Just because something looks pretty, or feels like a good idea, or has a lot of sunk cost in it, does not mean it should be pursued. Not even if it's generating revenue. The only efforts a new product team should be expending are those that lead to validated learning about customers. Here's hoping Steve will share those techniques with us someday. In the meantime, I hope some of you will find the lean startup a helpful framework.

Overall, here are the lessons I take from (the imaginary) Steve Jobs:
  • Hold your team to high standards, don't settle for products that don't meet the vision, iterate, iterate, iterate.
  • Be disciplined about which vision to pursue; choose products that have large markets.
  • Discover what's in customers' heads, and tackle problems where design is a differentiator.
  • Work on as few products as possible, keep resources in reserve for experimentation.
  • Start over (pivot) if you find yourself with a product that's not working.
As I said, applying these principles in a startup is different from a very high profile public company. I've tried to abstract them a little so we can examine them at a level where they might translate. So far, everything I see is compatible with what I believe. So thanks, Steve, for the inspiration, the great products, and the great advice. Here's hoping future innovators who will follow in your footsteps are reading today.

Time for me to sign off - my Macbook Air is really burning up my lap. Hey, Steve, seriously, this viable product is a little too minimal...
Reblog this post [with Zemanta]


  1. REALLY appreciate you posting this, Eric. Still working through the post, but I wanted to say that this clarifies the distinction well. I have some more questions, but I'll ping you with them a little later on. But, yeah: thanks for this.

  2. Eric,
    This isn't as strong as your usual posts. The "mapping" of the imaginary Steve Job's (imagined) process to "lean principles" is a bit strained (and too tenous, imho).

    "Steve Jobs doesn't put out crappy products and ask for feedback" seems valid to me as does the stark contrast that SJ's methods (whatever they are) seem to have with the Lean Startup ideas.

    Nothing wrong with that of course. Even if SJ has some unique "process" that enables him to come up with killer products (I suspect he is just wired differently from most people and probably has no real idea of how he does it to a point where it is teachable), that doesn't mean that the Lean Startup/Steve Blank's ideas etc have less of a value for the rest of us.

    In the end, SJ just might be the genius who makes his own rules, the exception that proves the rule. And that's fine.

    My key takeaway from your post (as I said, the mapping of the imaginary SJ's process to Lean Principles is a bit tenous, at least in my perception) is this idea..

    "Getting feedback is *not* about abandoning your vision"

    I see too many people expecting the customer to tell them what to make vs doing the "incorporate feedback into the original vision" part. The distinction seems subtle but is there nonetheless.

    Thanks for writing this. Made me think!

  3. I really enjoyed this article, my first time reading any of your stuff. Will start to pay attention...

  4. Great post Eric.

    Also SJ's obsession with better taste (positively) contributes to the overall product development process.

    Good taste and common sense goes a long way.

  5. Beyond Apple, we can find examples of the Steve Jobs character at Toyota, in the form of the Chief Engineer. The Chief Engineer, through many years of experience, has a "visceral feel for what the customer wants" and acts as "the voice of the customer".

    I jotted down some musings about the Chief Engineer role and software development a few months ago:

    In software, I tend to call this role "Product Director", but it has the same meaning and description. In talking with Mary Poppendieck about it, she said that the Chief Engineer role at 3M was called a "Product Champion". The name is less important than understanding the depth of competence that this person has, and what it takes to build it.

    There are some good anecdotes about Chief Engineers in the Toyota Product Development Book. Some of these reaffirm that they do indeed validate assumptions through interaction with customers. And in some cases, they disregard the results of customer focus groups if they sense that the results are incongruous.

    We might be a ways away from the reality of a Chief Engineer or Product Director role commonly in software development, but ultimately it's going to be a matter of getting the culture right so that people stick around long enough to develop not only the tactical skills, but also the honed instincts and senses.

    Thanks for another great article!

  6. Market Awareness, Visionary Innovation, Focus (discipline), Leadership, Execution

    Always appreciate your perspective Eric.

  7. I would not say that Apple "iterates, iterates, iterates." I'm not saying they should, or even that they can - just that they don't. Not compared to any other company of their size, at least. The lean startup (as I understand the concept) does not have major releases. Apple is all about big announcements.

    I agree with the point that SJ "just knows" what to do. But for every SJ, there are thousands of other executives who don't, and that's what makes lean startup so valuable.

  8. From a few facts and a strong commitment to your pre-existing notions you have conflated a version of Steve Jobs that may not exist at all. Not everything has to follow lean startup and customer development principles, and I write that as a devoted practitioner of both. It would be far more useful to discover how Apple really does develop their innovative products it rather than assume that they follow the principles that you and I espouse. It could be that their resources allow them to use methods that would be prohibitively expensive for startups. It could be that Jobs is just an irreplaceable genius blessed with unique insights and an ability to recruit, motivate, and retain other world class talent.

  9. Good post, Eric! I would recommend reading "Inside Steve's Brain"for solid insight on Jobs/Apple if you haven't read it.

  10. Eric, thanks for your insights.. a long-time Apple user and watcher, I think you have captured a good part of it. You make me think of something I've been pondering for a while but haven't been able to articulate: the realignment of company and "customer" in the age of pervasive social networks and innovation. Companies constantly gaffe by asking the wrong questions, they don't understand their role and the customer's role the way Apple does. The customer understands experience, and social networks, since they make pervasive p2p economically viable, accentuate customers' ability to mirror each other and articulate customer experience. But they have no clue about how to design, make, distribute and service products that enable them to have experience, that's the company's competence. As you've pointed out, Jobs totally gets that. Companies that abdicate to customers will fail, as will companies that don't listen and respect experience. Too many companies are obsessed about making money and Apple, a public company, proves that they can make money by not focusing on it. Making money should never be a focus, it's a result. Thanks again for some great insight. Cheers-

  11. I'm amazed at the level of engagement this post has generated - thanks to everyone who's commented.

    I know this post was unusually speculative, so I get why there are so many comments along those lines. Let me stress again that I don't claim any special knowledge of how Steve Jobs works, nor am I even trying to extrapolate from the available data.

    Instead, I engaged in a thought exercise designed to answer the (highly speculative) question I get asked quite often: is my approach compatible with Steve Jobs?

    I think the available evidence is pretty slim either way. But so many people have mythologized Jobs that I think the "irreplaceable genius" theory is just as speculative as anything I've proposed here.

    So I don't think the Apple experience should be used to discount lean, agile, or customer development-type approaches. I do hope that someday more information about the _real_ internal process comes out, so we can get some answers to our open questions:

    - how much internal (secret) iteration goes into each product?
    - how many products are "launched" internally or with beta testers that never see the light of day?
    - what is the mortality rate of prototypes or new ideas? Where do they come from?
    - what is the "right discipline to think through whether a lot of other people are going to want it, too?" Is it done by Jobs or does it have an institutional basis?

    In the meantime, I hope my speculation provokes some interesting thoughts.

  12. My guess isn't that Steve Job's just know what people want, rather he got lucky a few times, has help but gets the credit and we don't hear about all the failures he and his gang have had. It's like talking about the miraculous cures of "faith healers" at Lourdes but failing to ask them where are the crutches of those they haven't healed. Also, a lot of Apple's stuff seems to me to be mostly "eye candy" and like fashion jewelry it's trendy and sells. Maybe the real "secret" here is that the Apple store at Valley Fair is really just a department of Nordstrom's.

  13. Did you know Harvard is trying to trademark the phrase "lessons learned"?

  14. 1) RE: 'At Pixar...We stopped production for five months' - stopping production to fix problems is a thoroughly 'lean' concept as promoted by Taichi Ohno and implemented by Toyota in the form of 'andon' in engine assembly as early as 1950!!!!!

    2) Let's not forget that apple has has also made products that quite gone to plan

    Thanks Eric for the continued enlightenment, I thoroughly enjoyed your Lean Startup Webinar!

  15. There are a couple of videos that are relevant to this discussion. The videos are Steve explaining the company's path forward to Next employees.

    While the methods they used to gather the facts aren't obvious, hints are dropped along the way. For example at 8:47 into the first video Steve says: "Every single customer we've talked to ..."

    The videos look and sound a lot like a "Lesson's Learned" talk. He goes over in a incredibly lucid way who the customers are and how their product will satisfy their needs.

    Steve Jobs at Next part 1:

    Steve Jobs at Next part 2

  16. Phil, thanks for posting those videos - I've always liked them. Relentless focus on customers and their real problems seems like a common thread. And I do believe there is a fair amount of iteration happening behind closed doors.

  17. Nice commentary.
    There is however a very distinct item to remind yourself. Every single great Apple product on the market today is not something incredibly innovative.
    No new, ground breaking technologies in the iPhone. iPod was not the first one of it's kind.
    The main point is that Apple has this idea of making things easy to use. They start with existing technologies, and make them easy, not obstructive and good looking.
    Lack of true groundbreaking technologies, is the reason I stay away from their products. But otherwise they are really great.
    That is very important for a start up that is aiming at consumer market - it has to be easy to use and fashionable, is possible.

  18. Hi Eric, thanks a lot for the post.

    I'm reading "Inside Steve's Brain" book, and I highly reccommend to understand Jobs approach on launching new products.

    They do lots of prototypes internally.

    Tks a lot for sharing. It clarified a lot the lean startup method with Apple's approach.

    Best wishes,

    Miguel Cavalcanti, from Brazil

  19. I can't claim to be an authority on the process at Apple, but I did spend a year there in '93 (when Steve wasn't there).

    What I observed was that large numbers of projects were started internally, they were allowed to run for around nine months, then 90% of them would be culled, and the staff were moved onto the 10% of projects that remained.

    Obviously this approach is unsuitable for most if not all start-ups

  20. Maybe Steve Jobs just has a different definition of "viable"? What a normal person would view as acceptable, he would not. I do believe that Apple iterates and is one of the keys to their success. Apple's first hit product was Apple II, which came after I. The Mac came after Lisa, which was a flop. The iPhone came after many generations of the iPod. The first gen iPod was created using OS and GUI software from Pixo, which was a little start-up (I know because I was an investor and on the board of Pixo). Apple would NEVER license the OS of a product which was strategic and expected to be a huge hit! They (Rubinstein was the lead) later acquired full license to Pixo OS (before the company was acquired by Sun) and then even later replaced it with their own OS. Apple also hired away key engineers from Pixo, who I believe still work there.