The New Tech CEO Archetype

When your tech company was in need of new management you used to go get a proven executive, like Lou Gerstner or Meg Whitman, who had experience running large companies.

But now, it seems, you go get a strong technical person who rose up the ranks of product management and knows how to ship great products.

The new tech CEO archetype is a computer scientist who got into product management early in their career, led large product teams at a big important tech company, is in their 40s, and has great taste in technology, tech talent, and most of all tech products.

Marissa Mayer, Satya Nadella, and Sundar Pichai are examples of this archetype.

It’s not really different from what we look for in startup founders. Most of the time, the founders we back come from product backgrounds. They have a track record of building and shipping products. They are technical and can go toe to toe with their engineering team. They understand where technology is headed and they understand how software products are made and evolve.

When young people tell me they want to start or run a tech company, I always tell them to go work in product at a big tech company. I believe that product is the heart and soul of tech companies, it is where it all comes together. You can’t build a great company without great products (or great people).

So it’s heartening to me to see that the next generation of technology leaders is coming from product management. I think that bodes well for those companies and the tech industry in general.


Comments (Archived):

  1. awaldstein

    A great product is the fair advantage that every true entrepreneur should have.It doesn’t guarantee anything but it damn well gives them a shot.

    1. Mario Cantin

      Without it they can’t come up with a growth hack — there’ll be no liftoff.

      1. awaldstein

        Without it you have nada for certain.I”ve won without the best product. Different game that i wouldn’t play in today’s market.

        1. Mario Cantin

          I have a sense of where you’re coming from. Its a different world out there now for sure.

  2. LIAD

    Alongside the hard skills they also need to have a innate desire to create, a driving urge to delight and a burning passion to beautify.

    1. Mario Cantin

      Now the question is, could someone who has the innate desire to create, along with a driving urge to delight and a burning passion to beautify, but lacking in technical skills make up for the latter by having an over-supply of the former?

    2. fredwilson

      Yes. For sure

  3. steve nson

    Interesting point. My question is why do we need Archetypes to begin with? Each company is different, each Founder is unique and each CEO has their own strengths and weaknesses. When you start putting people in boxes it leads to “pattern recognition” which leads to not accounting for people of other backgrounds who could bring a different perspective to the team.The future great companies and leaders will not emerge from obvious backgrounds. We live on a planet with over 7 billion people, with diverse interests, goals, problems and aspirations and the best way tap into this genius is to accept each persons individuality and encourage them to bring their talents to the forefront.

    1. LIAD

      Building a co is formulaic. Starts with an awesome product. Tech companies levarage software as they’re core building block. A product centered CEO with a strong technical understanding, comfortable directing engineers is a force multiplier.

      1. steve nson

        Nothing is formulaic. If that were the case all these well funded startups wouldn’t fail.

        1. LIAD

          Nothing is guaranteed but you can set the board up optimally.

          1. steve nson

            What is the failure rate of well funded startups that had the right teams? There are so many factors that go into making a company, person or team succeed and it cannot just be boiled down to what their technical background is.The leaders motivation matters, their ability to handle adversity matters, their flexibility matters, their stubbornness matters, the team dynamics matter, and the list goes on. Their is no secret formula, you have to make the assessment based on the specific circumstance and the individual in question.

      2. James Ferguson @kWIQly

        I prefer the notion of starting with market need for a solution. An engineer sees solution and builds.That’s how bridges and rail ways got built.The ‘ingenuity’ in engineer means they may be social medical or traditional engineers but that’s what they do_ get stuff done.

        1. JLM

          .The training of the engineer mind is pure logic.Find/define the problem.Find/define the natural laws, equations, conditions that support the solution.Craft the solution.Go build it.It is very much like the legal mind set.When I was studying engineering, I had a prof (Retired Admiral) who required you to put a rectangle around your answer and to underline it twice.The class sizes were very small and you worked the problems at the blackboard every day.If you failed to do it, “F”.To this day, I remember the lesson — present your answer so that people know it is the freakin’ answer.JLMwww.themusingsofthebigredca…

          1. LE

            I suspect that the engineering mind, just like the legal mind, can also be very limiting and retarding to free and creative thought and problem solving. The parody regarding using a pencil instead of a pen in the space shuttle is an example. Another might be the engineers who couldn’t figure out (as the fictional story goes) how to unstuck a truck under a bridge when some kid yells out “just let air out of the tires”. No doubt exaggerations but they illustrate my point.With the legal mind I have seen cases where an exec with legal training frames everything as they have been taught as a legal issue and quite often has a hard time departing from that thinking (sure it’s possible but…). [1] Ditto for any particular strong educational underpinning or background. My point is simply there are clear negatives to thinking a certain way it’s not just “another tool in the box is always good” (after all more tools does mean more weight).Sounds like the prof (retired admiral) was somewhat rigid maybe that’s good on a nuclear submarine or a ship but it can be a disaster in many other places.Pure logic or pure legal logic is not the answer to many problems. Or is engineering. Not that you are saying that of course.[1] I am remembering a contract that was given to me for colocation space way back in 2002. The VP at the small colo space was an attorney. And predictably he had a complete overkill long and clusterfuck multi page contract that they wanted signed as if they were AT&T and had millions of customers. Point being statistically doing what they were doing it was not necessary given the market and the actual potential damages and most importantly the probability that any of their customers would actually hire an attorney and take the legal route (having to prove damages..)

          2. JLM

            .No doubt that every skill also limits our broader view of things.”When you are a hammer, everything looks like a nail.”OTOH, when I went to my 45th college reunion, all the engineers who had had the Admiral still remembered statics, dynamics, thermodynamics and heat transfer.He had a memorable style for unformed lumps of clay like us cadets.JLMwww.themusingsofthebigredca…

          3. LE

            From college, the person that I remember most was Stuart Samuels a film lecturer who consistently got the highest marks in the rating books at Penn. (Not given tenure however think the others were jealous of his popularity or film wasn’t important..)http://www.stuartsamuelspro…I remember very little from college but I do remember, when reviewing Hitchcocks Psycho Samuels saying about Janet Leigh “Hitckcock designed that Bra” in order to drive home the point about his obsessiveness with details.And my god, here is a picture of lacy undergarment:…

          4. Twain Twain

            Better to be a Swiss Army knife. We can’t open beer bottles with hammers, haha!

          5. JLM

            .You’re kidding right?Of course you can open a beer with a hammer. That’s what those two sharp things on the back are for. Right?JLMwww.themusingsofthebigredca…

          6. Twain Twain

            No, the prongs are for prying nails.

          7. JLM

            .Silly goose!Today everything is about multi-tasking. Those teeth are for pulling nails AND opening beer cans — can open a bottle in a pinch.JLMwww.themusingsofthebigredca…

          8. Twain Twain

            Haha, so then the handle is for what?

          9. James Ferguson @kWIQly

            Simply depends on how well you define the problem.Untested assumptions are your enemy

          10. James Ferguson @kWIQly

            Underlining _ AKA marketing

          11. Girish Mehta

            In his “Lost Interview”, Steve Jobs made the mindset analogy with a lawyer – at 0:23 secs. Now you said it.It had struck me when I had first heard it – he separates out becoming something (he says “I don’t think anybody should be a lawyer”..) from learning how to think in a certain way.

        2. Twain Twain

          To an extent I agree.However, one of the things about engineering purists (and this includes in code engineering as much as structural engineering) is that they think in terms of functional utility but not necessarily in terms of aesthetic appeal, comfort of use, how to charge for use of that bridge and what it stands for as inspiration through the tests of time.This is the distinction between the Product person and the pure Engineer. The Product person also problem-solves those other factors.

          1. James Ferguson @kWIQly

            Agree I guess I see things from the perspective of client desperation.We provide a solution to some knotty problems that often get wrapped up into our channel partners solutions so aesthetics mean less directly and bm is negotiated.However you are right Sometimes a little Socratic wisdom goes a long way Know what you are rubbish at and get help!I engineers did this they would perform better – Design ego is an issue.

          2. Twain Twain

            Absolutely. I’m rubbish at some things so I figure that if I hang out on AVC, sooner or later someone smarter than me will share something random that makes me go, “A-HA! THAT was the thing I didn’t know I didn’t know and now I know it!”

          3. James Ferguson @kWIQly

            That’s nigh on a Direct quote from Plato’s Apology maybe if Socrates had used your turn of phrase he could of avoided the hemlock altogether;)

          4. Vasudev Ram

            I read somewhere that apparently he didn’t really care to avoid the hemlock after seeing state of affairs and calibre of his opponents at the time.

          5. James Ferguson @kWIQly

            You are absolutely right.It was meant as a weak joke. I understood it was almost a stoic protest vote on the basis that his mates Plato included had offered to buy his freedom. The conclusion_ that he preferred not to fund the state corruption.Brave man brave decision and eloquent matyrdom for his stated intent to continue defending truth.

          6. ShanaC

            Yup. And also questions if it is even an engineering problem in the first place

          7. Vasudev Ram

            Excellent point.

          8. ShanaC

            why thank you

          9. Vasudev Ram

            >This is the distinction between the Product person and the pure Engineer.Good point, except that I would alter it a bit to say – the holistic thinker + doer, not (just) the Product person. Because such ability as you describe is not restricted to Product persons (those whose titles on an org chart include the word “Product”).Said another way, such a holistic person is the real Product person …

        3. sigmaalgebra

          My old view: Start with two lists, list (A) of technology means and list (B) of “market needs”. Then pick a pair, one from list (A) and one from list (B). Pick the best pair, that is, essentially the pair that promises to make the most money fairly quickly with very high probability.So, “market need” — a space elevator. Build the first one of those, or even really just know how to do that, and quickly retire and, if you wish, run for POTUS. Only problem: No one knows how to do that. So, the space elevator is a great item on list (B), but there’s nothing on list (A) to do it.Uh, what’s a space elevator? Put a heavy ball on a long, strong string, say, 25,000 miles long. Put the ball in a synchronous orbit at altitude of, say, 25,000 miles. Then, to put something else in orbit, just let it climb up the string. So, get a super cheap way to put mass into orbit. List (A)? Need a really strong string, stronger than anyone knows how to build.Another great item on list (B): A safe, effective, cheap one pill taken once to cure any cancer. A great item on list (B). Unfortunately, there’s nothing corresponding on list (A).So, can’t just pick a “market need” and, instead, have to pick a pair.

          1. Twain Twain

            Well, of course, if there’s nothing corresponding on list (A) some people may be of the disposition to…INVENT IT.For example, let’s track back to the time before mobile phones rather than to the non-sensical space elevator example.There was a market need for a mobile phone; people wanted to be able to communicate, on the move rather than tied to the landline telephone on their desk.Some of the technology needed was not on list (A) so some people invented it.

          2. sigmaalgebra

            Well, of course, if there’s nothing corresponding on list (A) some people may be of the disposition to…INVENT IT. Yup. I’ve done such invention several times. Some such were just academic research publications, but some were directly useful in practice.And I did such invention again for my current project. So, I have a pair, a “market need” from list (B) and, from my invention, a corresponding “technology means” from list (A). And now, running software.In the world of high tech, there is a lot of interest in “market needs”. But, alas, there’s exactly one person in the world who cares about my “technology means”, me.In high tech, there’s essentially no interest at all in good, new, powerful, valuable “technology means”. E.g., no investor will evaluate it.Ah, I’ll just call that situation, i.e., that so many people ignore new items on list (A) important for big “market needs” on list (B), as an advantage for my project.Really, we’re still back with the Mother Goose story “The Little Red Hen”.That lesson was a surprise, but by now I’ve learned it, thoroughly. Time to move on.

      3. Richard

        Read slowly….There are two camps. Some people are so bright that they remarkably need little hands on tech experience to manage…others are so bright that they need years of hands on tech experience to manage

        1. pointsnfigures

          If you see a problem, and are NOT a techie-you need to be able to find the right partner and articulate your solution so they can build it. This is tougher to do.

    2. Richard

      Tech companies are nothing new. From big pharma to big farm equipment, we’ve been in the tech biz for 100 yearsIn many ways many of today’s tech companies sucesses or failures seem little to do with the technology.

      1. Twain Twain

        The successes and failures are all to do with…PEOPLE.The tech is simply an extension / tool invention of those people.

        1. Richard

          Lots of successes out there that are so not withstanding the people

          1. Twain Twain

            People doesn’t necessarily just mean the founders and team.It also means the customers who see the tech as an extension / tool invention for them.

          2. Twain Twain

            That makes me think of Zuckerberg’s views on Twitter, by the way.Now, detractors of Twitter could say, “Of course it would have been successful without the people it has, anyway!”Umm…no…Anyone remember Myspace or a million and one other IM platforms?

          3. creative group

            Twain Twain:those IM platforms failed because of the lack of execution. The management in those platforms stunk.

        2. LE

          Observing business for such a long time as I have I believe “all” is a bit to strong.Success has to do with timing, luck, serendipity and a host of other factors. Additionally the saying “a rising tide lifts all boats” (which is really just a version of luck) is appropriate especially with tech.

          1. Twain Twain

            Haha, great point about timing!The other day @wmoug:disqus emailed me a great framework for MVPs and I pointed out that it needed 2 additional factors: TIMELY and scalable.The reason the people factor matters is because people form the market for that product. Sometimes, a startup can do all the market research in the world, engineer a really useful product and it will still not fly because people didn’t feel they could relate to it.Blockchain+Bitcoin have some of those challenges.

      2. Dave Pinsen

        I’ve always thought it’s sort of odd for “tech” to apply solely to Internet companies, as if Exxon Mobil, for example, weren’t a tech company.

        1. Richard

          Or a MRI co, you got to like private equity, they could care less about being cool.

    3. Twain Twain

      Fred’s post focuses more on the common PROCESSES these particular CEOs went through and learned from, honing their craft, rather than on their education, personalities and other diverse factors.The thing about tech is the same as the thing about industry and banking: to rise to the very top (aka CEO), it’s better the person’s worked closely with Product.That means they’ve been through every part of how that product is made (including materials sourcing), distributed, marketed, scaled and sold.

    4. John McGrath

      I agree people should be judged on their merits, but pattern recognition is a real thing and more than just confirmation bias. When a VC evaluates a company, or you evaluate someone you might work with, of course you should start with the particulars–but also consider their background. If someone was a senior PM at Google that tells me something.Some future great companies–maybe the best of them–will come out of left field. But imo most probably will come out of obvious backgrounds.

      1. steve nson

        My view is that tech should be the great equalizer, where you went to school, where you worked or who you know should not be as important as what you understand about the problem you are solving and your potential customers. I understand the need of having some key indicators that you use to determine a good vs bad investment but I would rely more on the other factors I mentioned above instead of whether someone comes from a product background. Understanding sales is as important as any other skill a good CEO should have.

        1. creative group

          Our unscientific observation is that East Coast people are more concerned with where you attended school and the West Coast people it is about ideas and products that make money first.

    5. Laura Yecies

      Great point – I completely agree that the danger of “pattern recognition” is promotion of bias.

    6. Vasudev Ram

      >My question is why do we need Archetypes to begin with? >When you start putting people in boxes it leads to “pattern recognition”Got to agree. I think this is like machine learning – let’s leave it to the machines, they’re being programmed to do it anyway – and use our uniquely human talents to do what they cannot – judgement, taste, differentiation, etc. (Said as a programmer, BTW).

    7. creative group

      Steve:your perspective was not only refreshing but shows that thoughts, information and sound judgement can evolve from other than an elite group who chooses who becomes the next King or Queen. (The lack of diversity in the tech sector of women, minorities, etc highlights the facts verses what the talking points being disseminated).

      1. steve nson

        Thank you Creative Group.

  4. Anne Libby

    a) Working in a big company also gives you an idea of how a “machine” that works, works at scale. (for better and for worse.)b) The idea of archetypes brings *Moneyball* to mind…a fantastic book about the assumptions we make when we constructing teams.

    1. creative group

      Billy Bean never won a championship with that formula.

      1. Anne Libby

        I don’t think that was the actual point. I think the point was to run a more profitable ball team.

        1. creative group

          Anne:Professional baseball for 80% of Owners, Executives, Managers and players is about winning which results into profitability. The more a team wins the more fans they will garner, apparel selling, etc.But we did look at the point of an owner in Oakland not concerned with winning but money and the team and fans suffered that awful experiment. Didn’t work out to well. It introduced analytics to sports as a tool and not the platform. Analytics can be thrown out the door if it was used the last four years. Ask the opponents of the San Francisco Giants. Inferior teams that won the World series on paper verses the opponents they beat. Understanding baseball assists in that assessment.Thanks for your view anyway.

  5. Shalabh

    Is having a strong technical background necessary for becoming a PM or can a non-techie, who otherwise understands technology quite well, learn along the way?

    1. ErikSchwartz

      Define “otherwise understands technology”.To me the technical basis a PM needs is an understanding of complexity theory.O(1) vs O(n) vs O(n^2) vs O(log n) vs O(n log n) and so on…

      1. ShanaC


    2. fredwilson

      The latter can work well

  6. William Mougayar

    I can’t totally agree. The fact that this is happening doesn’t mean it is necessarily a good thing for these companies. What seems to be a weakness for the Archetype you have described is a lack of appreciation or understanding of marketing, and that’s a serious handicap for these companies.A good CEO is one that is well rounded in product, technology, management, marketing, and leadership skills. Steve Jobs comes to mind.

    1. Mario Cantin

      Steve Jobs is the quintessential tech company CEO, as well as an outlier case.Growth hacking is on the rise and it is disrupting traditional marketing, no?Looking at the valuations of the more conspicuous examples of companies that have catapulted themselves upward that way, it would seem to me that there is ample marketing taking place, it’s just done differently.You might aslo be referring to startups who fail to gain traction, but I doubt traditional marketing is what will make up for the difference.It definitely looks to me like it is in fact about finding the product/market fit and coming up with a growth hack. And when you get really big, like Yahoo, Google, et al, then the “hack” might very well become hiring the big-gun MBA’s.

      1. Girish Mehta

        Re “Growth hacking is on the rise and it is disrupting traditional marketing, no ?”Respect your point of view, but I tend to disagree with this. Growth hacking has its place, but it doesn’t either disrupt or fill in for marketing. My sense is that over the next 12-18 months this will be increasingly recognized in the startup space. Thanks.” A business has only two basic functions – Marketing and Innovation” – Peter Drucker.

        1. Mario Cantin

          I would bet that if one of us is wrong, it would more likely be me.Makes sense, a hack is well… a hack. It’s not standard procedure; and as you’ve alluded to, it has its limitations.What I wonder, though, is how much traditional marketing will have to morph in response to the advent of the “new” economy.

          1. Mario Cantin

            Very nice William, thank you. I’ve added it to Pocket so I consume the content later.

          2. sigmaalgebra

            Kept it! Thanks.

        2. William Mougayar


        3. JLM

          .Who is Peter Drucker?What does he know about it anyway?JLMwww.themusingsofthebigredca…

      2. William Mougayar

        Steve Jobs himself was an outlier, but what he did is what each CEO should do- including being a great communicator.Equating growth hacking to marketing is wrong, and I’ve written about this exact issue this week – this isn’t traditional marketing. It’s just marketing. http://startupmanagement.or

        1. Mario Cantin

          Timely discussion then :-)I’ve added that one too to my reading list.

      3. Eleanor Haas

        “Growth hacking” has always been an integral aspect of marketing. We just now have more ways to do it – SEO, websites, etc. But content marketing is nothing new, nor is A/B testing. What’s been disrupted is reliance on one form of marketing communications, i.e., advertising.

        1. Mario Cantin

          Interesting perspective.

        2. JLM

          .Well, only one half.One half of advertising is brilliant and the other is crap.Our problem is figuring out which half is which.JLMwww.themusingsofthebigredca…

    2. Bernard Desarnauts

      Yep. Other examples include Marc Benioff – a one off/superb marketeer among other things, Larry Ellison, any better sales and tech guy out there?

    3. Twain Twain

      Steve Jobs was a Product person not an engineer / coder, as Steve Wozniak has confirmed in his interviews.”Steve didn’t ever code. He wasn’t an engineer and he didn’t do any original design, but he was technical enough to alter and change and add to other designs. I did all of the Apple I and Apple ][ myself, including the feature choices. I did all of the BASIC myself (it’s in handwriting as I couldn’t afford an assembler). The only person who helped write some of the Apple ][ code was Allen Baum, who helped with the ‘monitor’ program.” — source:…Now, even in the field of PMs, there’s a kaleidoscope of types that span from Steve Jobs to Marissa Mayer.Jobs famously said, ” It’s really hard to design products by focus groups. A lot of times, people don’t know what they want until you show it to them” and would quote Ford too, “If I had asked people what they wanted, they would have said faster horses.”Clearly, they weren’t into market research which is a process staple of PMs who are in the Marissa Mayer mould. She’s famous for AB market research testing every single feature on a web page, including 41 potential shades of blue for the Google search box and its pixel width.Of the two, Steve Jobs was arguably the better marketer despite the apparent lack of market research.

      1. William Mougayar

        That’s OK. Engineering or product background are both good. It’s about how the person evolves & develops in running a growing company.Steve had an eye for design and usability too.

        1. Twain Twain

          In the earliest formation, the founder(s) have to be hustlers, guerillas, rebels and definitely grafters.As growth happens and, especially when interacting with institutional investors and corporate clients, a much more polished and nuanced strategy (even patrician) is needed.That’s often why CEOs are replaced at later stages. The people who are hustlers and code nerds may not fit the mould that institutional investors are familiar and comfortable with.Some CEOs like Steve Jobs have that rare magic quality; staying rebel whilst also being relateable to Wall Street-types.

    4. Richard

      someone who knows how to code : find someone who started at age 10. Someone who knows how to market: find simeone who read adage at age 10

      1. PhilipSugar

        Find somebody that did both.

    5. LE

      Steve Jobs comes to mindWould also add said person, as Jobs did, has to have tremendous charisma.Without that you will never be able to connive and convince others.That’s a big reason why Donald Trump is doing so well. It’s more than the money and the outrageous statements. As much as he turns some people off he has the “it” factor that makes people want to listen to him and believe him and follow him. I am not talking about those expecting the train wreck and entertainment either (that’s an entirely different issue). In one of his speeches he gave examples of military generals such as Patton and MacArthur in that respect comparing them to what we have today.

      1. William Mougayar

        Yup. I can’t think of a leader who moved big mountains who wasn’t charismatic or charming.

        1. JLM

          .Read about Troy H Middleton, the Corps Commander who made the decision to hold Bastogne in the Battle of the Bulge.http://themusingsofthebigre…When he told Eisenhower he could hold it, Ike told his entire staff, “Troy says he can hold Bastogne.”That was all it took to create the entire strategy to defeat what could have been an incredible defeat for the allies. The word of one man.He had such force of character that his word was like steel. His character was the defining force of his achievements in the military (youngest Regimental commander in WWI, 3-star in WWII, and the man who made Bastogne work) and civilian life (President of LSU who built it into a university with graduate schools).The 10th Armored Div was in Bastogne and Middleton made an incredibly prescient decision — he asked for the 101st Abn Div because he knew paratroopers were used to fighting while surrounded and he had correctly surmised that the Germans would surround Bastogne.Both MacArthur and Patton ran great staffs. Eisenhower, as a young officer, served under MacArthur on his staff.While they were both charismatic, Patton moreso than MacArthur, it was because they had superb staff work shoring up all of their decisions.Patton began to turn two of his divisions toward Bastogne before he knew how bad the situation was. When Ike asked him long before he could come to the rescue of Bastogne, he said 48 hours and he could attack right from his movement formations.Montgomery had answered the same question by saying, “Two weeks.”Patton was quite the military scientist and logistician. Had he had the gas supply that went to Montgomery, he would likely have taken Berlin.I tell you this because there is a lot of substance behind very charismatic leaders and sometimes there are guys like Middleton who were quiet, unassuming killers. The charismatic guys get all the press.JLMwww.themusingsofthebigredca…

          1. LE

            Curious on if you could comment on my theory that wars are much harder in the modern age simply because we aren’t as able to have any civilian casualties with any of our actions. Every decision has to be perfect prior to shooting or bombing and my thought is that while this is great for those who end up alive it’s ultimately bad for us in terms of winning. [1] In the glory days coverups were no doubt easier and possible to pull off. That gave those leading a great deal of latitude don’t you think? Not to take away from their greatness of course.[1] Not to mention we are going to see the same thing with police in our country or even with respects to government and privacy. Inability to accept some casualty for the greater goals.

          2. JLM

            .The onset of asymmetrical warfare — wherein we are not fighting sovereign nations but movements — has changed things completely.We should view wars as a process of destroying our adversaries, scattering them to the winds, destroying the little infrastructure they have, and robbing them of their access to capital.Where we make our mistake is the unique American notion of rebuilding “countries” we defeat or conduct military operations within.Future wars should be incredibly violent — like the first 48 hours of the first Gulf War when we sent Baghdad back 300 years — and short.Get in. Kill our enemies. Destroy their infrastructure. Communicate to the world what we did and offer any like minded shitheads a slice of the same pie.This approach will minimize friendly casualties, be much cheaper, and will require regular maintenance.The specter of civilian casualties is an element of warfare that has always impacted America more than any other country in the world.It is part of our national character and it is a good thing up until it hamstrings a Lieutenant trying to decide whether to return fire while his men are under fire.The ROE (rules of engagement) are always a problem.JLMwww.themusingsofthebigredca…

          3. Girish Mehta

            Curious how General McAuliffe and the “Nuts” story (Dec 22, 1944 surrender demand of Bastogne) relates to this…i.e. how McAuliffe and Middleton are connected. I did not know about Middleton, but I know the “Nuts” story, so intrigued :-).…Colonel Gaston Bell: General McAuliffe refused a German surrender demand. You know what he said?General George S. Patton: What?Colonel Gaston Bell: “Nuts!”Patton: [laughing] Keep them moving, colonel. A man that eloquent has to be saved.- Francis Ford Coppola and Edmund H. North, in Patton (1970)

          4. JLM

            .Troy H Middleton was the Corps Commander while Gen McAuliffe was the ADC (assistant division commander) of the 101st Abn Div, one of the divisions in Middleton’s Corps.The Div CO (commanding officer), MG Maxwell Taylor, was in the US at a conference and when they got surrounded was not able to get back into Bastogne until much later.McAuliffe fought the division.It was one of the greatest lines in the history of warfare. It was defiant, bit unrealistic as the Germans had a stranglehold on the 101st, the 10th Armored, and Bastogne.He was, apparently, very well known for never using profanity which in the paratroopers would have been a rarity.Today is National Airborne Day — not a holiday that too many folks celebrate — and this comment typifies the Airborne spirit.JLMwww.themusingsofthebigredca…

          5. Girish Mehta

            Thanks…yes, apparently the General’s vocabulary was atypical.

          6. James Ferguson @kWIQly

            JLM. THAT is a great story.The idea that keeping wits when taking hits from all sides would make a difference is a fantastic example of emotional intelligence probably not often associated with warfare. BUT so incisive

          7. JLM

            .If you want another example of a guy who could use his personality to drive combat outcomes read: Chesty by Col Jon T Hoffman USMCR.Chesty Puller — VMI man — fought in Nicarauga, Santo Domingo, Guadacanal, Cape Gloucester, Peleliu, Okinawa, Inchon, Wonsan, Chosin Reservoir and then the Punchbowl Area.He commanded First Marines at the end of WWII and Korea — a very unusual arrangement.[Many folks have a hard time understanding the Marine regimental system. “First Marines” means the First Marine Regiment, not the First Marine Division. First Marines is one of the three maneuver regiments in the First Marine Division.]He got a bad rap for Peleliu in which First Marines was wrecked by the end of the campaign. The navy prep for the landing beaches failed to hit a huge promontory on the left side of Chesty’s beach which he had to take on with infantry.Every man I have ever spoke to about their experience on Peleliu says the same thing, “We took the Japs out with a spoon.”Chesty got the First Marines again after the NKoreans invaded the south and a short time thereafter they were at sea headed to Inchon. He never even had a chance to train them up but many — most –were WWII veterans and it was a battle tested bunch of leaders.They landed at Inchon and took Seoul in a week. The Marines did not have urban warfare experience in WWII so Seoul was a new kind of fighting.In the return from the Chosin, after the Chinese entered the war with almost half a million soldiers, First Marines had the rearguard for 1st Mar Div and fought their way out, often outnumbered by 20:1. The retreat from the Chosin was one of the greatest feats of arms in the history of warfare.The Div Cdr, OP Smith, was brilliant building supply dumps while headed north plus an air field to evacuate wounded.Chesty Puller had that same quiet confidence.In the Marines Corps they used to every night drink a toast to: “Chesty Puller, wherever the Hell you are.”His son was horribly wounded and lost all his limbs as a 2nd Lt USMC in Viet Nam.A very good read as a testament to the power of personal example and character for a leader in the toughest possible conditions.JLMwww.themusingsofthebigredca…

          8. Twain Twain

            See? This is the reason you’re one of my sensei.As a teenager I read military bios (Boxer Rebellion, Spanish Armada, WWII) but this completely tailed off.Now, I read your comments and have added a book to my reading list.

          9. ACR

            Just reading this as it happens:”The personnel of this army of Sherman had names that meant nothing beyond its picket lines… At the War Department or in the White House at Washington, as an instance, the name of Joseph Anthony Mower meant nothing in particular, a rather ordinary division or corps commander. Yet Mower was a phenomenon, one of the strangest personalities of the war. No West Pointer, a Mexican War volunteer private, in September of ’61 commissioned a captain, he so fought and marched with his men that in August of ’64 he was made major general of volunteers. A rough, irascible, hairy six-footer, swearing through his tawny dark whiskers, ‘you always had to look for him in the front line,’ said other officers. ‘He never spoke of himself,’ said Sherman, who added without reserve, ‘A better soldier or a braver man never lived.’ … The reports were monotonous: ‘General Mower drove the enemy two miles.’ He wrote few letters, seemed careless of distinction. Several successive sets of his staff officers had been killed in action while trying to keep pace with him. … on Sherman’s request Mower had been ordered East, had hurried to join up for the march from Savannah. In the forefront, a terrific marcher of men, he led them through rain and mud, through icy hip-deep waters, taking the punishments of bad weather and outdoor sleeping with his troops. His fellow officers could hardly find words to describe how superbly he handled his men in the push through the Salkehatchie swamps. One of the greatest of American soldiers, he was also one of the many valorous, picturesque unknowns…”from Carl Sandburg’s Lincoln biography

          10. JLM

            .Good story. However, it is very difficult for me to read anything related to the first American war criminal, Wm T Sherman.JLMwww.themusingsoftthebigredc…

        2. awaldstein

          you think either of these words are appropriate to trump?

          1. JLM

            .Part of Trump’s appeal is simply the blandness of some of the other candidates. More than half remind me of a boiled potato.In the land of the blind, the One Eyed Jack reigns supreme.Part of his “charisma” is folks see him doing battle with the media and contesting them on-the-spot.Three weeks ago, I was predicting, with some certainty and conviction, that Trump was 1.4327 gaffes from destruction.Now, I am not so sure.I will tell you this, the freakin’ Republican establishment has no idea how to deal with him. No clue.Their guy is not just failing to hold serve, he is falling.This is going to be a very interesting election.Everyone seems to forget how well the Republicans did in 2014. There is a very good chance that 2016 is just a Ground Hog Day for the Republicans.JLMwww.themusigsofthebigredcar…

          2. Vasudev Ram

            Hindi saying: Andho mein kana raja.Meaning: Among the blind, the one-eyed man is king.

          3. JLM

            .I wonder if all cultures don’t have the same basic sayings.I am going to start pretending I am so worldly that I can make Hindi aphorisms.Like I said — “In the land of the blind, the one-eyed Jack is King.”Thank you and bless you.JLMwww.themusingsofthebigredca…

          4. Vasudev Ram

            Cool 🙂

    6. JamesHRH

      A good CEO is maniacal about the customer experience and sets the vision accordingly (unless the company is an asset play).Everything else is variable, as long as there as it meets the baseline of effective management (lots of type of ice cream).

      1. William Mougayar

        It’s not one thing more than the other, as it’s tough to generalize. A CEO’s job must be the most rounded. That’s why they are CEO’s.

  7. Twain Twain

    This is the first post of Fred’s where I wholeheartedly agree!Product is the intersection and apex of creative, business, strategic and technical skills, imo.What I’ve noticed are MBAs who’ve never coded before or done any product design who try to sell themselves as Product Managers.NO. A true PM knows how to code. They’ve likely been coding since they were kids. Along the way, they acquired business knowhow which made them understand how to wrangle the code to make products that market would PAY FOR and love.It takes 10,000 hours to become great PMs like Mayer, Pichai and Nadella.I have a few thousand more hours before I can be like them. In my case, my first Product experience was when I was 17 working for the world’s second largest aromachemicals company.You have to apply science and techniques to invent and make consumer products that ship and sell.Later in startups and in banking, that product knowhow gained further training.It’s why I can design and code my systems and figure out what the business strategy and customer utility are.Tens of thousands of hours of practice, doing and constant learning…….

    1. fredwilson

      The first???

      1. Twain Twain

        Indeedy. On 99.99999999% of your posts there’s something I feel I either need to learn more about, disagree with or would do differently.The entirety of the post may make sense but I’ll spot that bit of information where I go, “Ok, wait a sec…Let’s drill down on the context and map the associations with what he’s just shared…Hmmn, the crosshairs are on target but it’s just fractionally not quite a perfect fit for me.”This post, though, is perfect.I especially agree with “can go toe to toe with their engineering team. They understand where technology is headed and they understand how software products are made and evolve.”So… my first-ever presentation at UBS was to the Global Head of eCommerce. Based on a single product wireframe, he signed off “carte blanche” on my platform idea. He didn’t even ask for a costing; just told me to get on with it so I did.The Director of Engineering had 15+ years of banking technology experience and was not the type to “suffer fools gladly” (like most engineers). I had about 1 month of investment banking experience and 6 months from the Wealth Mgmt business where we’d rolled out a pan-European brokerage. I was a good decade younger than everyone else.Still, he and his team agreed to build my e-Intelligence platform for ZERO cost and it was way ahead of schedule because they believed in the product and I’d scoped the technical roadmap so thoroughly it was a win+win “no brainer” to code and plug into existing technologies.Even today, I go toe to toe with engineers I know (PhDs in AI, Stanford / Cambridge / Imperial grads, teenagers who’ve won Maths Olympiads) and they don’t give me an easier time just because I’m a woman.If anything, they give me a harder time because their initial bias is, “She looks like a kooky Boho artist. Probably knows jack about programming.” Maybe that’s why I started to dress in sneakers and hoodies to fit in, haha.[I did once experience an engineer assuming I couldn’t code. In fact, he tried to screw the team over by saying that, without his engineering skills, we couldn’t do anything. So I re-coded the app that had taken him 16 hours to do in under 3 hours. Then I kicked him off the team for lack of team values and consideration.]Product people need to be “Big Picture” about why people will love and benefit from what’s being built as well as SYSTEMATIC about how each part of the stack gets made and is efficient+effective.Engineers will press PMs on obscure Haskell or library scripts or database routing and a good Product person needs to be able to identify why and how whatever the engineer is proposing does / doesn’t improve the user’s experience and the bottomline of how much that technology costs compared with the rates of ROI.I’ve worked with people who are phenomenal engineers (few could wrangle Objective-C, Python or firmware like them). However, they’re terrible at Product because they can’t get beyond their own nerd-clique use case for it and they have very narrow, limited experience of accounting and strategy.PMs also have to have great people skills and be able to flow between being tactful and being brutally honest when a product isn’t working or will take time to generate profits.There’s a massive difference between Product CEOs and CEOs who’ve risen through the other business functions (sales & marketing; finance; engineering; design).Product CEOs tend to be vertically and horizontally skilled.That integrated perspective is how the great ones know where technology’s heading, so they can put all the product pucks in place before everyone else in the market.

        1. Richard

          Interesting, this is the first post that seemed completely out of left field.This blog never seemed to be about running a fortune 500 company nor learning from a Fortune 500 companies.Seemed like something a private equity money manager at Kkr might be focused on.

          1. Twain Twain

            Haha, well maybe Fred decided to switch up his play a bit — could be he watched some sports game and noticed all the action was on the left flank!Joking aside…The Product is front & center of startups and Fortune 500 companies alike, imo.In business school, we learn about the 7Ps of the Marketing Mix.Being a “left of field” type because I’d worked in industry before I arrived at b-school, frameworks such as the 7Ps seemed a bit off-target.After working in small startups (6-50 people) as well as behemoths like UBS and IFF, I noticed that Product is the thing the other factors coalesce around.

          2. Twain Twain

            @disqus_Awy3Cl8ObF:disqus – Product.

        2. ShanaC

          See, I wonder if that description of engineering culture is a fault of pms giving into the clique culture of engineering.

          1. Twain Twain

            I’ve been coding since I was a pre-teen. In the 20+ years I’ve been in and around (mostly male) developers, the culture (which is often intimidating for women) hasn’t changed much.It’s been fine for me, though.With the exception of that one engineer, most guys have been helpful and they appreciate I’m focused on inventing and shipping products.

      2. Richard

        I thought there was a typo there as well!

    2. ShanaC

      How well do pms need to know how to code? Why as well as an engineer?

      1. Twain Twain

        A couple of years ago, there was a heated debate in a PM group in NYC.Some PMs were completely resistant to any form of coding. They’d been business consultants and weren’t particularly interested in learning how to code. Their reasoning was, “Engineers will build what WE tell them to.”Meanwhile, some of the other PMs had started to learn Ruby-on-Rails. They said they knew they’d never be as good as the engineers who’ve been coding for decades but they still wanted to know enough to do their jobs better.Then there were PMs who’d previously been Lead Engineer (had MSc Comp. Sci too) and migrated over towards the business side as their managerial responsibilities increased.So different PMs have different views on how much code to know.

  8. Dan Storms

    This is a recognition that the user has taken more control of the buying decision. Traditionally the business manager picked the feature list that they liked best. Now a lot more products are evangelized by users. So knowing what makes a great product is essential to optimizing a bigger company.

  9. Steven Kuyan

    I think this comes from a desire by the technical co-founders to lead the vision of the company. It’s hard to make decisions on a company when when your immersed in the day to day coding. I’ve seen a shift of technical co-founders wanting to write code to wanting to lead the companies so this seems like a natural progression.A technical co-founder has an edge when the company is in its early stages but they need to learn more than just product if they want to be a tech CEO. I’ve seen far too many opt for shipping product instead of hiring, fundraising, etc and to support the contradicting arguement, there are a number that I know that can’t wait to stop writing code. Working in a big company or a growing startup can certainly provide the learning experience for a developer to become a CEO, but there are many other factors that contribute to a solid tech CEO.I am glad that many of them are making that decision, education for CS degrees should contain parts of an MBA/Product Management to help catalyze this. We’re trying to do something similar at NYU for CS/Management students, hopefully the results will be positive.

    1. William Mougayar

      Well said. An over emphasis on product and engineering will have blind spots, especially after a startup gets a lot of users. I’ve been saying this a lot lately.

  10. WA

    I hope there is the understanding that their employees and management team are not products. In the name of an esprit de corps and not the Amazon approach of Workplace Darwinism. Timely Topic.…This said I like that the importance of the team and its people is stressed here often – forgive the plug of a disturbing article…

  11. pointsnfigures

    CEO to me depends on stage of company. I don’t think you necessarily have to have an engineering background, but it does help at early stages. When the company grows, it’s about managing people.

    1. JLM

      .Agreeing more with you than you do with yourself.Well played.It is all about the size of the company at an instant in time.JLMwww.themusingsofthebigredca…

  12. William Mougayar

    Having a great product and vision is where it starts. It’s absolutely necessary.But taking the product to the market is even more important. That’s what differentiates companies and CEOs.

  13. Tom Labus

    It Nadella gets MSFT a second life, he goes into the Hall. It looks like it too.

    1. Twain Twain

      I switched from Windows to iOS about 8 years ago and haven’t been much interested in MSFT since. Certainly not ever interested in coding for .NET frameworks.However, I saw this marketing for Windows 10 and found myself thinking, “Wow, their vision’s amazing. Right now, Apple’s doing ‘If it’s not an iPhone, it’s not an iPhone’ strategy which is about protecting a cash cow, Google’s busy re-arranging its Alphabet and MSFT may be the one that’s LEADING in vision of innovation terms.”*

      1. JLM

        .Fabulous bit of video.JLMwww.themusingsofthebigredca…

        1. Twain Twain

          It’s the “a more human way to do” tagline that puts a stamp on where MSFT strategy’s going under Nadella.Compare this with Google’s Alphabet announcement:”We are excited about…* Getting more ambitious things done.* Taking the long-term view.* Empowering great entrepreneurs and companies to flourish.* Investing at the scale of the opportunities and resources we see.* Improving the transparency and oversight of what we’re doing.* Making Google even better through greater focus.”* And hopefully… as a result of all this, improving the lives of as many people as we can.”That sounds like a laundry list by management consultants for bankers rather than speaking to the hearts of consumers which is what MSFT does in that video.

          1. JLM

            .It is the classic storytelling dilemma — show, don’t tell.Google will do just fine but they may not inspire people.We are in that silly season in the US — politically — in which we are looking for inspiration not just a good resume.JLMwww.themusingsofthebigredca…

      2. Richard

        Wow! Steve Cook is the least appreciated CEO in history. Apple have executed essentially flawlessly.

        1. Twain Twain

          I’m a fan of Tim Cook. There are few operators of his calibre.Whether he’s a product visionary, though, remains to be seen.There are certainly pieces of tech Apple has which, if they were integrated into future releases, would get developers excited all over again.Interestingly, of the big techcos, Apple went from being ahead in AI with its SIRI acquisition (and its almost immediate implementation into the OS) to Apple hardly featuring in AI articles or threads within the AI communityIn fact, whilst Google, Facebook, Baidu, Amazon have been building up their AI teams in the last 2-3 years and having AI researchers “Ooohing and ahhing”, Apple only recently posted vacancies for just over 100 roles.Maybe the strategic partnership with IBM — and specifically IBM Watson — will help Apple with the AI piece they’re currently behind in.It’s not just Apple’s “This is where we are and the great products we’ve executed” that count. It’s also, “These are our product INSPIRATIONS and ambitions for the future” which developers, consumers and investors alike are interested in.

          1. Richard

            What’s the source for Apples… googles AI hiring?

          2. Twain Twain

            LinkedIn on Apple AI. In June+July they posted over 100 roles which is now at 47:*…This number of roles indicates they’re building up a new unit.Google has made the most AI-related acquihires and been most vocal on their interests in the space:*…* http://www.businessinsider….Facebook started building an AI unit in NY in 2013:*…*…Facebook acquired Wit.AI which has 6000+ developers registered on it, to solve some Natural Language problems:* http://uk.businessinsider.c…Baidu’s AI unit wants to hire 150-200 AI people by end 2015:*

      3. sigmaalgebra

        Sorry, I’ve seen that video before and similar visions well before that.IMHO, the whole thing is totally, laughably wrong in vision, intention, conception, and, no doubt, especially execution.My conclusion was that Windows 10 will be the serious mistake that ruins Nadella.For me, Windows 7 and/or Windows Server.It’s still the case: Computers just do what they are told to do. Further, when we tell a computer to do something, we have something fairly definite in mind. That in any reasonably general sense a computer will guess what I want and do it for me is absurd.For that vision, have to be smoking funny stuff.It’s an old story, that somehow computers can think — not in any very real or useful sense, not yet or for a long time. We are closer to curing cancer, and that because we have a good shot at knowing fine detail just how cancer works. Still, we are years away from curing cancer. For intelligence, as needed for the artificial intelligence visions, we don’t have a clue about how that works, not in a human, kitty cat, puppy dog, or even an octopus.That there are books, courses, and job titles in artificial intelligence means essentially nothing but absurd wishes and outrageous hype.When I was at IBM’s Watson lab working in artificial intelligence, I gave a paper at AAAI IAAI (“innovative applications”) conference at Stanford, and all the good work at that conference was just good, traditional engineering.There is more good engineering to be done; so far there’s nothing in thinking computers to be done.So, net we are farther away from anything like thinking computers than we are from curing cancer. Again, for a thinking computer, we just don’t know even where to start.

        1. Twain Twain

          Oh so then you won’t be impressed with Google founders’ vision for AI either (http://www.businessinsider…..

          1. sigmaalgebra

            Right. I’ve written software that can reason, do things that would require a human to think to do them, and do some very narrowly defined things in a very narrow context better than humans can. If permit such hype, then could say such things about a mechanical calculator. But intelligence? Not a chance.I know, we get into the Turing test, etc. But just watch, say, a kitty cat learn — sorry, Page and Brin, no one knows how to program that, not even with your favorite collection of C++ classes.

          2. Twain Twain

            Well, then, you be bemused by this claim that the machines are already more intelligent than us:* http://www.technologyreview…Of course, this is if you believe that IQ tests are in any way a complete and reliable way of measuring intelligence in the first place…

          3. sigmaalgebra

            Even taking the claim at face value, they just engineered a narrow solution to a narrow problem and are still a very long way from intelligence. Of course, this is if you believe that IQ tests are in any way a complete and reliable way of measuring intelligence in the first place… Give a test of 500 questions. Reduce the data to principle components (factor analysis). The component with the largest eigenvalue, define that as intelligence quotient.So, for one number it is the best linear predictor on all 500 questions. Done. That’s IQ.

          4. Twain Twain

            Now, I happen not to believe that Human Intelligence should be defined or measured by existing tools (IQ tests, personality tests etc.).The reason is because existing tools miss out on and can’t measure other parts of our intelligence. Such as: * bodily-kinesthetic experiences (of the type that sports people would have in which they learn how to throw a ball better, for example);* smell intelligence (of the type flavorists and perfume “Noses”) have; and* oral linguistic intelligence.In tandem with this, we need to consider the Quantum Energies expended by our human intelligence activities.In any case, as you know well, mathematicians like to IDENTIFY THE PROBLEM and DEFINE THE FUNCTIONAL VARIABLES.So…I made a start…Now, this is not to say it’s better than all the existing models for intelligence, logic, reasoning etc. like IQ tests and maths exams.It’s simply a mathematical hypothesis from which products (systems) can be built — just as Google, FB, Amazon, Baidu etc all have their hypothesis, borrow maths from the ages and build systems to test their hypothesis.[And the language wouldn’t be in C++.]

          5. Roger Toennis

            Hey Twain,I just came back around to read your reply post to me in that “carrying two phones” comments thread last month. (Re: X-Code, Y-Code for male female archetype modeling.)I want to see more on this.Also We need to talk directly. See this series or articles my fiance’ and I are writing about the Masculine/Feminine Archetype using a new lexicon we are framing we call “Versatility Factor” or “V-Factor”.…I’m rtoennis AT Gmail.www.linkedin/in/rogertoennis

          6. sigmaalgebra

            Ah, come ON!As I outlined above, for IQ just ask 500 questions that you believe have something to do with whatever you regard as intelligence, do a factor analysis, that is, first cut, just find the principle components, take the eigenvector with the largest eigenvalue, and use that to define IQ.I will omit the linear algebra details — I’ve been through them too many times and don’t want to do it again. The key, of course, is the polar decomposition. In the nonsingular case, there’s an easy proof; in the general case the problem is not that there is no solution but there are infinitely many solutions and just have to pick one.In that way, get the best, single, linear approximation to the answers of all 500 questions. Look it up: It’s about two pages of linear algebra.For a long time, that stuff looked to a lot of people as an important evaluation of people. Some of the history is ugly.If course if you have a dozen questions, say, about your ” bodily-kinesthetic experiences” or “perfume”, that have answers independent of the 500, then of course the IQ you get will also be independent of the dozen answers, that is, in particular, with correlation 0 and, as a linear predictor, useless. For a clean proof of such things about independence, see the treatment of independence based on sigma algebras, e.g., in J. Neveu.That’s about all there is to IQ.You are considering thinking ability, power, capability, etc. in a much more general sense. Then, of course simple, 100 year old IQ doesn’t measure what you want. So, don’t struggle with IQ. So, for what you want, just f’get about IQ!Simple, right?

          7. Twain Twain

            That’s exactly the point!!!I am considering thinking ability, power, capability in a much more general rather than the narrow IQ test sense!!!I don’t give a fig about following in the linear algebra path of IQ tests, actually.

          8. sigmaalgebra

            The classic linear algebra used in anIQ test could be applied to whatyou are interested in, but I doubt such simple, linear math would bevery effective. That is, you’d justget poor measurements.

        2. Twain Twain

          There are three separate issues in your reasoning.(1.) NEUROSCIENCE================As I’ve quoted several times before: “When we talk mathematics, we may be discussing a secondary language built on the primary language of the nervous system” — John Von Neumann.As I have also written several times before, the tools we currently have can’t tell us how the brain works precisely.For laymen, we’ve been presented with simple ideas for what the left+right sides of the brains do (see images).For AI researchers, existing maths is the best tool we have for modeling how the brain works and is how we build AI.You may have missed this comment of mine the other day when a Professor of AI wrote this: “For many years, researchers working in AI struggled with the challenges of making a machine “think” like a human. Programs were written to play chess, which became powerful enough to beat many of the world’s champions. But the programs often relied upon brute-force computing power. True intelligence – artificial or otherwise – always seemed elusive.Recent advances in neural networks have changed all that. A neural network is software that mimics how the human brain works. It “learns” how to solve a problem by comparing many possible solutions, and then improves the solution with more data. These programs learn very rapidly, and quite soon they can solve problems much larger and more complicated than a human could.”To that, my response was: “A neural network does NOT mimic how the human brain works. It is a mathematical method of applying Könisberg’s bridge problem with end-objectives and constraints (shortest time required, crossing over fewest number of bridges or links to get to a node) and finding optimal pathways.A neural network is in no way like-for-like to human intelligence and our organic synapses which carry with them biochemical and perceptual factors (e.g. emotions, language, culture) that affect our intelligence, decision-making, experience memories, how we compare options and do risk management and actions that thinking as our behaviors.Frankly, presenting machine neural networks as exactly equivalent to human neural networks is nonsense.It’s akin to saying a cog’s the same as a combustion engine. Or flapping’s all we need to fly!”So I am with you about how the existing models are far from true Thinking Machines.There are also these articles:*…*…(2.) THINKING COMPUTERS======================To date, our frameworks for such machines have assumed that Mathematics measures Intelligence. Even in the language reasoning part of standard IQ tests there is a SCORE.In Natural Language Processing there is:* Markov Model…* Google Word2Vec Model…* Stanford Recursive NN…Maths as a language has been generalized over to framework language as a language.WHETHER THIS IS VALID…IS AS YET UNPROVEN (and you know how mathematicians love proofs and universal application!).(3.) MSFT VIDEO ADVERTORIAL========================It starts not claiming they can execute on all of it yet but…”Imagine….(i.) These kids won’t have to remember passwords…(ii.) Or obsess about security….(iii.) For them, every screen is meant to be touched…(iv.) And Web pages are meant to be scribbled on and shared…(v.) They’ll expect their devices to listen to them and talk and sing…(vi.) Tell a funny joke…(vii.) And as they grow and get better at things, technology will too…(viii.) They’ll do things that parents never even dreamed of…(ix.) Because these kids grew up with Windows 10…(x.) The future starts here. For all of us. Get started today.Let’s break these down and see if the technology already enables this and if MSFT has a good chance of executing on it:(i.) Depending on us setting Password Pre-fill on our devices, it’s true we don’t need to remember passwords.Apple has TouchID which means we don’t have to remember it either and there are other biometric methods being researched by MSFT (facial recognition, eye scan).(ii.) MSFT like all the major techcos makes security a priority.(iii.) MSFT devices are indeed touchable; see Surface tablet as well its earlier iteration from 8 years ago —*…(iv.) There are a number of Javascript and HTML5 plugins that do enable scribbling on web pages for sharing. Wikipedia is an example of a collaborative canvas.(v.) Cortana does indeed listen and talk and sing.(vi.) There is MS research on Natural Language and humor.(vii.) The ability for the machines to do unsupervised learning is being implemented by MSFT and all the major techcos.(viii.) We can’t currently do what MSFT envisages for 2019:*…(ix.) Windows 10 is the start of this strategy — again see MSFT vision for 2019 video for context.(x.) MSFT call to action.We can debate the technical challenges of formulating and developing suitable frameworks (mathematical and other) for AI that more accurately reflect our natural intelligence.In terms of strategy and marketing communication, this MSFT video has hit its mark as inspiration and ambition backed up by their existing execution.In comparison with MSFT advertising in years past, it has character, sets the mood and is hopeful instead of merely functional and dry.

          1. sigmaalgebra

            Recent advances in neural networks have changed all that. A neural network is software that mimics how the human brain works. Horse feathers. The same old hype. E.g., use neural. Maybe with that could emulate one neuron of a worm, but even for that I’ve seen nothing although it may have been done. Keep that up and can get a little of a worm.But a worm’s not so stupid: It can detect food, move to it, make baby worms, avoid predators, etc.And we don’t have a clue how to go from an emulation of one neuron to the rest of the capabilities of a worm. And software for playing chess, checkers, space invaders, etc. is not closer. It “learns” how to solve a problem by comparing many possible solutions, and then improves the solution with more data. These programs learn very rapidly, and quite soon they can solve problems much larger and more complicated than a human could.” Hype. Deliberately misleading.For “learns”, can say that about dynamic programming and/or stochastic optimal control, Markov decision processes, etc. from R. Bellman, E. Dynkin, R. Rockafellar, or me — I wrote my Ph.D. dissertation in that, and there’s nothing intelligent about it. Instead, the computer is just doing what it was quite carefully told to do in a quite limited context. Useful? In some cases, yes. Intelligent? Nope.”Learns”? Can say that about my software, really the math it implements, for my project. Useful? I believe so. Intelligent? Nope. So I am with you about how the existing models are far from true Thinking Machines. Good. WHETHER THIS IS VALID…IS AS YET UNPROVEN (and you know how mathematicians love proofs and universal application!). Right.What a brain is doing has to be mathematically something, but so far we don’t have hardly a clue about what. It is a mathematical method of applying Könisberg’s bridge problem with end-objectives and constraints (shortest time required, crossing over fewest number of bridges or links to get to a node) and finding optimal pathways. Optimization, i.e., mathematical programming, e.g., as in the journal Mathematical Programming, is a nicely developed field. Parts of the field are quite old going back to, say, Newton and sliding a bead down a wire. Parts of the field are relatively new from operations research. It’s some applied math, often quite nicely done. I studied some of the best of it in grad school. Part of it is, say, the Kuhn-Tucker conditions for nonlinear programming. While I was in grad school, I wrote a paper on that — solved some long outstanding problems. Later I published the paper in Journal of Optimization Theory and Applications (JOTA), a decently good journal.Optimization is intelligent? Not a chance. Frankly, presenting machine neural networks as exactly equivalent to human neural networks is nonsense.It’s akin to saying a cog’s the same as a combustion engine. Right. To date, our frameworks for such machines have assumed that Mathematics measures Intelligence. Well, thinking has to be mathematically something, but we don’t have even as much as a weak little hollow hint of a tiny clue about what.A fundamental problem with natural language processing is that it is input to thinking, and we don’t know what that is. (3.) MSFT VIDEO ADVERTORIALIt starts not claiming they can execute on all of it yet but… In my response, I had no patience with the low level details of that video clip.E.g., for Let’s break these down and see if the technology already enables this and if MSFT has a good chance of executing on it:(i.) Depending on us setting Password Pre-fill on our devices, it’s true we don’t need to remember passwords. I don’t try to remember passwords now. I generate new passwords easily, frequently — just type in some 8-15 or so single digits in a really sloppy, random way. Firefox remembers some passwords for me, and the rest I have a super easy way to remember and make use of the rest based on, right, just my favorite editor, KEDIT. Actually, after I’ve created a password, I never type it again but use one of several editor macros to put the password on the system clipboard so that I can do a paste operation to get the password in some single line text box as needed.Here might be a password:password29047839203Good luck trying to crack that.So, in Windows 10, could Microsoft do some things with passwords? Sure. We’ve known a lot about passwords, authentication, encryption, etc. back to Kerberos, RSA, etc. for a long time. And Microsoft has made heavy use of Kerberos and RSA.Sorry, Nadella, that password problem you are trying to solve, I don’t have. Moreover, for more than I am currently using, e.g., as will be needed by my startup, you have some solid, old products that do pretty well now, e.g., Active Directory Domain Services as in…right, based heavily on Kerberos and encryption.But again, I have no patience with such low level, simple details of that clip.Instead I responded to the suggestion in the clip that Windows 10 would be intelligent, and that’s just hype.Then, for the remark in that clip about security, what a hoot: Microsoft’s record on security has been abominable — until just a few years ago, they were still fixing just simple, old buffer overflow errors. Buffer overflows are errors in Programming 101.I have to conclude that mostly Microsoft has been laughing at computer security, just releasing buggy software and laughing at the harm it does people and how they have to keep coming back to Microsoft for fixes, one bug at a time.Nearly each time Microsoft comes out with new software, it also comes out with a huge new collection of security vulnerabilities.For Windows 10, some of what it is doing now, e.g., sending user data back to Microsoft, looks like an open door to new, grotesque security problems.And when Windows 10 sends my data back to Microsoft, then 10 is supposed, with the help of the cloud, to help me get my work done. Nope. E.g., for writing a technical paper, Microsoft still doesn’t have anything better than Knuth’s TeX; this means that I will still be using TeX; then, with TeX, no way will Microsoft be able help me. Not a chance. My favorite text editor is KEDIT — it is also my favorite tool — and Microsoft has nothing even 10% as good. So, 10 won’t be helping me use KEDIT.My view is that Nadella blew it. Maybe he will make enough money to retire, and then Gates will have to look for another bright guy to “take Microsoft to the next level”.It’s a waste of time to debunk outrageous hype. It’s even worse to be taken in by it.

    2. fredwilson


    3. Richard

      I can see it taking on the role of it. I don’t see it as a consumer Play.

  14. hemang c subramanian

    Nice observation. However the case of Marissa is very different from either that or Sundar or Satya. Marissa was an external hire at Yahoo! while both Sundar and Satya were long time insiders. Sundar having spent more than a decade and Satya having spent more than 2 decades with their respective organizations. They grew up the ranks navigating organizational structures- but for Marissa she was hired at a time where no internal candidate could fill the role effectively.The great thing however is that capital markets, employees and partners have reacted positively to these appointments.

    1. fredwilson

      Good point

  15. JLM

    .It is difficult to make generalizations about CEOs when you consider the state of product, the size of the company, and the nature of the market. Still, Fred’s view is correct.If you had a chance to command a row boat, a landing craft, a minesweeper, a coastal attack vessel, a destroyer, a frigate or an aircraft carrier — you would likely jump at the chance to command an aircraft carrier.But, if you were “qualified” to do just that, you would likely have commanded one or more of those smaller vessels.You would have learned your core business — product — and you would be ready to stretch yourself to the next higher command.The military constantly feeds this progression — you go to your branch basic course, a company commander’s charm school, your advanced course, Command & Gen’l Staff College (tactics), and the War College (strategy, national security).There are variations on the theme with advanced schools that actually convey degrees (School of Advanced Military Studies), graduate degrees at civilian universities in national security subjects (Gen Petraeus had a PhD from Princeton), and being sent to the schools of allied nations (high level networking to boot).I belabor this point only to suggest that learning “product” is not a novel approach.Neither is a career path that is more likely to produce a high level leader and manager.The big thing in the tech world is the differentiation between entrepreneurial endeavors in which the enterprise is both driven by and held hostage to the individual genius of one person.An entrepreneur is a person who would rather command a row boat of their own than be the #2 on the Queen Elizabeth.JLMwww.themusingsofthebigredca…

    1. Tom Labus

      It’s always suprising who steps up in the heat of battle.

      1. JLM

        .You will find there are men who have almost no other redeeming quality or signal — as if they were stitched together for that one purpose and nothing else. They were lying in wait for that one opportunity and their tuning fork was calibrated just for that.You will also find there are men whose entire life has been nothing but excellence — classroom, athletics, extracurricular — who when the shit hits the fan are unable to lead. Not unwilling, just missing that gear.The common denominator of men who are recognized for valor is that they converted fear into action. When others failed to act, they did. They were often angry men whose anger could not be contained and who channeled it into action thereby inspiring/shaming their compatriots to act.My platoon sergeant won the DSC–not when I was his platoon leader. He was the calmest, nicest guy in the world until he wasn’t and then he was a tiger. Then he acted and reverted to the same guy.I asked him about it one night in the second half of a rum bottle with our feet next to a campfire — he almost didn’t know what I was talking about. It was organic.I know there are no extraordinary men — really — there are only ordinary men who rise to the occasion in extraordinary circumstances.JLMwww.themusingsofthebigredca…

        1. Tom Labus

          I’m reading about Hamilton creating the Treasury dept. Our debt was 30% of the economy. Amazing job and hated for doing it too

          1. JLM

            .Hamilton is one of my favorites. People forget he was also a very good soldier. Those were real men.Every time I am in DC, I walk past Treasury and look at his statue. Usually after a good meal across the street @ The Old Ebbitt.JLMwww.themusingsofthebigredca…

  16. Eleanor Haas

    It seems to me a trend rather than an archetype – the product person coming to the top of an existing organization – and it’s a logical outcome of rapid technological change and fierce competition.

  17. ThatOtherOtherGuy

    Marissa Mayer has not yet proven any real ability outside of her time at Google. Satya Nadella and Sundar Pichai are brand new in their role. I wouldn’t describe this as a new CEO archetype until these three actually accomplish something.

    1. JLM

      .Agree with the broad strokes of your comment. In the end, only results count.Bit of grit –Marissa Mayer was brought in to shake up Yahoo when the stock price was at $15. It closed at about $36 yesterday and has been as high as $52.She had a tough job. She has made some good moves and she has been criticized for some of her management style, actions.Still, she has delivered on the value proposition as measured by the stock price.She is a brilliant woman with two degrees from Stanford whose core expertise was thought to be AI.She made the leap to gen’l management.Fair play to her. She’s delivered.JLMwww.themusingsofthebigredca…

      1. ErikSchwartz

        YHOO is being propped up by investments in outside companies made before Mayer was CEO. The business that she actually manages is not doing well.

        1. JLM

          .May be perfectly true. The CEO gets the credit and the blame. She is not perfect. Nonetheless, the stock price is what it is.You want fair?Go to Dallas in September. Texas State Fair!JLMwww.themusingsofthebigredca…

        2. LE

          Yep Alibaba is the horn of plenty no question about that. [1][1] As opposed to the Horn of Africa where the Somali pirates attack merchant ships.

        3. Richard

          Yahoo feels like the jc penny of the tech era.There not done, but no one knows why.

      2. JamesHRH

        You don’t whiff much, but you are getting led astray but the financial headline.Name the unique anything that is Yahoo, other than a legacy audience.Erik’s comments below are bang on.

        1. JLM

          .You are introducing notions I did not put into evidence — I only commented on the share price because that was the yardstick they were holding when they gave her the job.The Alibaba saga is part of the story but she was holding the tiller when they crossed the Cape of Good Hope.Life is not fair. Go to Dallas in September. Texas State Fair — eat a Fletcher’s corn dog, a religious experience.JLMwww.themusingsofthebigredca…

          1. JamesHRH

            Well, Fred introduced the CEO Archetype, of which she is a poor example.She’s not a leader, she has no vision, she is a manager – set the course for me and I will execute bloodlessly – with a massive ego….the type that takes herself to the pay window on a short term basis and flushes the company down the toilet.Admittedly, others can come to the pay window with her, but they tend to be Wall St types, not people that she is charged with leading into the future.

          2. JLM

            .James, I am having a hard time understanding where you stand on the lady. Can you come out of your shell a bit more?Ooops, is that my sarcasm meter going off?JLMwww.themusingsofthebigredca…

          3. JamesHRH

            Well, you are a man with a delicate nature. I should be more genteel.

        2. ShanaC

          Isn’t that valuable by itself? And the fact that they keep it, and grow it?

      3. creative group

        To be fair Marissa hadn’t been giving a fair evaluation. If she was a man would she receive more room? Hard to tell. The stock price has mainly been boosted by the investment in Ali Baba which she had nothing to do with. Her predecessor made that smart move that couldn’t be realized until now.She has made acquisitions of companies to get the talent and then shutter those acquisitions. I will not call that brilliant.

    2. JamesHRH

      I agree completely.Rather surprised Fred chose them. There is a very good argument that Marissa is completely clueless.

      1. sigmaalgebra

        Naw! You are so cruel! How can you be so cruel?It’s totally clear: Well rested, nicely fixed up, her hair washed, etc., in one of her nice dresses, red or blue, she can look really cute, almost like still in high school. When she first joined Yahoo, she could look like she was still in middle school! We’re talking cute!Maybe her personality is some mixture of a great white shark, a rattlesnake, salt water alligator, and a Komodo dragon — a long way from sweet, meek, darling, adorable, and precious, like a kitty cat or puppy dog — but in one of her dresses with a good smile, she’s cute! Really cute!Sure, I’m a 100% American male and a total sucker for a pretty face, especially blond, small nose, dress size 4 or 6, ….Knowing what the heck to do with Yahoo? Not a clue! Not even a weak little hollow hint of a tiny clue!A CEO? Of what? Why think of her as a CEO?

  18. laurie kalmanson

    healthcare businesses have some parallels: being clinically brilliant and saving lives is vastly different from managing risk and infrastructure — at the top of the organization, both are important.

    1. ShanaC

      Which is why house can’t run a hospital.

      1. laurie kalmanson


  19. JLM

    .BTW, today is National Airborne Day. The day upon which America focuses its attention on those who have brought death from the skies on behalf of the nation.Thought you would like to know.One of your colleagues had 92 jumps and no broken bones. I won’t tell you who. It would make him blush. One time he almost got flattened by an artillery piece when its parachute was a streamer. Luckily, he was quick on his feet and dodged it.Of course, you did get an extra $60/month for doing it. Coupled with base pay of $277 plus hazardous duty pay of $60 and demo pay of $60 — it was a bloody fortune and no income tax when receiving hazardous duty pay.Those were the days.JLMwww.themusingsofthebigredca…

  20. brian piercy

    Awesome. There’s still hope.

  21. marko calvo-cruz

    Fred, would you blanket that advice to all young entrepreneurs? I think that there are some people who are just meant to jump in the fast lane. The obvious prodigies Mark Zuckerberg, Steve Jobs (yes, I know he briefly worked at Atari), Bill Gates, Matt Mullenweg, Shawn Fanning, even Evan Spiegel, (and I would argue Ben Casnocha, who is now Reid Hoffman’s chief of staff for those unfamiliar) come to mind though I’m sure I’m missing a bunch. It’s hard to remember that these people were ordinary folks before they made a name for themselves.Also, the notion that these people who were meant for the fast lane SHOULD be able to pull themselves by their own bootstraps largely divides this group into two: those raised in fortunate circumstances with families who have the connections and/or money to realize these kids’ potential (Snapchat, Microsoft, Amazon) AND those who lack these resources (inner-city kids turned famous rappers come to mind). But nobody, not even Steve Jobs made a name for themselves solely on their own merits. Some just have better inherit connections than others.(I believe that being friends with someone like Steve Wozniak who was ALREADY working on revolutionary projects is the salesman equivalent of a talented actress like Emma Watson making her name by getting cast as Hermoine in Harry Potter at such a young age).The latter group meanwhile, gets neglected by those who have the resources because it’s too difficult to find the diamonds in the rough- though that’s where the money is IF somebody takes a chance on them (Airbnb founders anyone?).This is probably the secret to Y Combinator’s success. They are willing to take the risk on these folk.

    1. fredwilson

      I always say if you have an idea that eats at you keeps you awake at night, then go for itIf not go work in a startup or a big tech company in product management

    2. steve nson

      Amen. I couldn’t have said it better myself.

  22. Sandy

    Product and technicals are really of secondary importance. That’s really the same thing as creating supply. Creating supply is just a skill that not only can be learned, but more importantly – it MUST be learned.Understanding human nature and psychology is far more important. That’s where user wants and likes come from. That’s really the same thing as seeing market demand where it doesn’t exist yet.Being able to see what is not there is not a skill. That requires natural talent, but that talent must also be honed by life experience. Ben Horowitz is exactly right that great CEOs are not born or made, they are found.The most important element of any company is always the element you can’t buy, teach, or train. If people can be one of the best just by putting in 10,000 hours, then that’s a skill, not a talent. There is a world of difference between the two.

  23. Candyman

    Amen to the part about getting some real operational experience where the rubber hits the road. I know that it’s become fashionable to think of experience as being unnecessary… but not with my money… 😉

  24. Ana Milicevic

    It’s not just about going toe to toe w/ engineering – but also w/ sales, finance, marketing and all other functional areas in the org. Having built and led product teams at just about every size and stage of company now, I’ve always thought of my role and that of product leaders in general as orchestra conductors – everyone in the orchestra knows how to play their own instrument (and perhaps a few others) exceptionally well; this holds true in a good tech team too. But orchestras still need conductors to be able to make beautiful music together. It’s really not a stretch from there to CEO (or COO for that matter) and I’m happy to see this emerge as a trend.

  25. ShanaC

    Basically, taste matters in all things.And there are different types of taste for different types of products.

  26. @DisruptiveDave

    Here’s another perspective from the early-stage-startups trenches. Far too often I bump into startups founded by “tech” and “product” guys who have this great skill to build things, but then they run into the “if you build it, they will come” hurdle. That one usually does more to kill a lot of these startups than poor product management does. (Again, I’m speaking from experience with early-stage, not-exactly-unicorn-status companies, for context.)

  27. sigmaalgebra

    Lou Gerstner or Meg Whitman Could either of them even set up a lemonade stand for less than $10 million?So, they are not likely to be CEOs of a bootstrapped startup.I don’t know much about Whitman — maybe at some point she will do some good things.But Gerstner was no bargain: He was sitting there with OS/2, at the time, the best operating system on Intel chips, said “I’m not going to throw a billion dollars into another PC operating system.”.First mistake: It didn’t have to cost a billion dollars. E.g., ask the people at MIT Project MAC that did Multics, DEC that did the DEC 10, Bell Labs that did Unix, UC Berkeley that did BSD, Torvalds that did Linux, Prime that did Primos, etc.He had some of the best people in operating systems to that time, e.g., George Radin “Three times in my career I tried to help IBM in operating systems, and three times I broke my pick trying.”. IBM didn’t want help in operating systems.When Gerstner made his statement, he had versions of OS/2 that would support multiple processors and run on the Mach kernel. OS/2 had the best communications stack in computing, and IBM had TCP/IP on a chip. IBM was running the Internet — the whole thing. IBM was the unique world experts in virtual machine — he could have been VMWare. DB/2 was running on OS/2 — he could have beaten Oracle, Sybase, SQL Server, etc. OS/2 was way ahead of Windows 95 and could have stayed ahead of Windows, 98, NT, 2000, XP, Vista, 7, … 10.Second mistake: Given the opportunity, i.e., the current Microsoft, a billion dollars would be cheap.IBM had everything in computing from X-ray lithography, processors, disk drives, disk subsystems, operating systems, LAN products, communications software, data base software, applications software, marketing, etc.Yet Gerstner blew it, totally blew it. He thought that the computing for the future was the computing he knew from the past. At the time, IBM was awash in brilliant people who wrote very well done reports that told him better, but he just didn’t want to listen. Net, he was a doofus, a very costly doofus.E.g., IBM Fellow Robert H. Dennard did Dennard scaling which explained why microprocessor line widths would be going down, down, down and why IBM was pursuing X-ray lithography. That all computer processors would be microprocessors was well known by Gerstner’s time. However, Gerstner just didn’t get it.As at… The manager does things right; the leader does the right thing. Maybe Gerstner did some things right, but it was Moore, Gates, Torvalds, Ellison, Page, Brin, Zuck, etc. who did the right things.

  28. sigmaalgebra

    So, a startup founder arrives at a venture capital firm with a startup with attributes for a huge market, with a good monetization plan, with traction significant and growing rapidly, including revenue and earnings, and is ready to sign a standard term sheet, and the venture firm will refuse to offer a term sheet due to an archetype?Or, without the attributes but with the archetype, the venture firm will offer a term sheet?It strikes me that what is relevant are the attributes, and I’m having a tough time seeing just how an archetype, or really much of anything about the founder, could be relevant to getting a term sheet.Net, the attributes are essentially necessary and sufficient for an offer of a term sheet so that essentially everything else, including archetype, has to be irrelevant.For Marissa Mayer, Satya Nadella, and Sundar Pichai are examples of this archetype.Is any one of these three an example of a successful startup founder? All three of them have worked for others who were the founders.For each of those three, ask “What have they done that was really good just on their own?”.First-cut, the guy who want to manage the company is the guy who successfully had the vision and made it real. Right? Or, while I’ve been typing code, the world has changed? Somehow, I don’t think so.

  29. Shripriya

    Could not agree with this more. One of the things we told PMs at eBay is “You are the CEO of the product”. It is great training to become the CEO of the company.

  30. sachmo

    Product management is pretty different than engineering. Most people (not all) that I meet in product management are decidedly non-technical. I think *real* engineers that can also work as product managers / R & D leaders / Quality Control Managers – that can sort of jump between roles as needed are the best future CEOs. On all 3 of the names above, the jury is still out on how they will ultimately perform. Let’s see how they do. This could be remembered as the great big mistake of letting the former product managers run companies…

  31. Asim Aslam

    The timing of this piece could not be any more fitting. Looked to start a tech company but was perhaps not ready. No longer in the mindset of an engineer but not quite yet the right time to be a CEO either. Product and strategy are the direction I’m headed in. Thank you for validating what I already felt in my gut.

  32. James T

    Product Management is also where big companies put acquired CEOs

  33. Mani Gandham

    First of all, what is this whole “tech” CEO / company thing? There’s no real definition for that. Most companies that SV considers boring use 100x more technology than the average webapp startup could dream of.Let’s get beyond the silly jargon first.

  34. LE

    Is this anecdotal or do you have data on this?The data you could say is the cumulative buildup of years of experience in doing what Fred does. And here is the thing about years of experience. The “rights” for someone smart who learns from their mistakes greatly outnumber the wrongs (based on solely your gut) and tend to cancel those out to an acceptable level. So in the long run using this as a shortcut for decision making, especially as your gut grows, tends to work more than it doesn’t.

  35. fredwilson

    Working in product at a startup is also good but at the start of their career it helps to work somewhere where there is training and strong management

  36. Twain Twain

    THIS.Early in a career, training and strong management instils a discipline for process that enables problem-solving.People like to romanticize startups as about chaos, gungho-ness and “flying by seats of pants fueled by Ramen”.Actually, startups are subject to processes in the same way as bigcos — just that the processes involve fewer people, less resources and they iterate at different velocities.They have to have a product pipeline, a strategy to market, the ability to show+tell, the knowhow to instruct lawyers (debt/equity structures, employment contracts), deal with and file accounts, book payments, negotiate contracts etc — all processes that affect startups and big businesses alike.

  37. JamesHRH

    I am with Charlie…..more than ever, understanding the market trumps understanding the tech.The tech is vital, but not the thin edge of the wedge. Name a technical UX that would blow people away right now….hard to do.Name a brand promise that could blow people away right now…..those are the innovations that are available late in major innovation cycle.

  38. Richard

    This is why Apple is the largest company in the world.

  39. JamesHRH

    Yes, they orchestrated existing tech into a new market category.Jobs was never a product guy, in the true meaning of the term (understood it, didn’t live it).

  40. Richard

    Yep, this is a special skill. The skill of not needing to drill to the bottom of the knowledge hole to know when to change direction.

  41. ShanaC

    YupSometimes startups miss this