Keep The Degree Of Difficulty Down

In many sports, like diving, gymnastics, skating, etc, the way to win is to perfectly execute a high degree of difficulty move.

In startups, I advise founders to avoid that way of thinking and try to execute a simple dive and hit the water perfectly.

I have sat through numerous pitches where I am listening to the founder explain their technology and go to market plan and I think “this is going to be a reverse triple somersault with two twists in pike and there is no way they are going to land it.”

There is plenty of risk in doing a startup. People quit their good paying jobs and take equity in lieu of cash, investors risk capital, customers are asked to try something new that might not work. Amplifying all of that risk with a high degree of difficulty product roadmap or go to market plan is crazy.

Unlike sports like diving or skating, you don’t have just one or two of three attempts to win. In startups, you get to show your stuff every day, all the time.

So the better approach is to pick something simple to execute, nail it, then build on it with another relatively simple move, nail that too, and keep going.

When, ten years later, you look back at what you and the team accomplished, it may well look like a reverse triple somersault with two twists in pike, and it will have been exactly that, and you will have won the prize too. But you will have done it by doing the easier things perfectly thousands of times instead of the hard thing just once.


Comments (Archived):

  1. Pointsandfigures

    Coaches I had growing up used to call it the KISS system. It stood for “Keep it simple stupid.” When you keep things simple, it’s easier to execute.

  2. Jason T

    So true. And it’s the hardest thing for many first time founders to understand.It is also the beauty of a minimum viable product approach as it frees the founders from overly-complex product decisions, and places the emphasis on product-market fit instead.

  3. William Mougayar

    In other words, don’t try to boil the ocean from Day 1.

    1. awaldstein

      Once you bring ‘market’ into any discussion, you have an ‘ocean’. That’s the core of this issue.

      1. William Mougayar

        but it depends on the approach taken. if you believe in doing it gradually, it is more credible.

  4. awaldstein

    Beautifully written Fred.Yes I agree.Big ideas, simple executions step at a time. Need to say that these twists often come from the other side of the table, squeezing go-to-market and projections on ideas that are powerful but still inchoate in many respects.Great VCs see this. Greatness is never the norm.

    1. fredwilson

      Thanks Arnold. I enjoyed writing it. This is one of those posts where I woke up with this in my head and all I had to do was sit down and write and it all came out in about 2-3mins. I love it when writing is like that. Rarely is

      1. Girish Mehta

        Its a important point but have a strong feeling of deja vu. Did you write a similar post (with the same comparison of degree of difficulty in diving/gymnastics) in the past several months ? Worth a re-write regardless.Maybe I am imagining stuff….

        1. fredwilson

          Anything is possible. I’ve written so many posts.

          1. Girish Mehta

            Ah, got it. I think I was remembering the point from the 1994 Berkshire Hathaway Chairman’s letter to shareholders.Thought I was losing my mind (ok, I still am…but thats a different matter..)http://www.berkshirehathawa…”Investors should remember that their scorecard is not computed using Olympic-diving methods: Degree-of-difficulty doesn’t count. If you are right about a business whose value is largely dependent on a single key factor that is both easy to understand and enduring, the payoff is the same as if you had correctly analyzed an investment alternative characterized by many constantly shifting and complex variables”. – Buffett, Berkshire Hathaway 1994 letter to shareholders.

      2. JamesHRH

        The interesting thing is that I am not sure Arnold has it right – so many massive entities are founded by someone without a ‘big idea’.They just aimed for what you laid out – this thing. Then the next thing.The one piece of advice I wished I had gotten @ 20.

        1. Vasudev Ram

          Yes, likely many of the old school entrepreneurs for example. Many of them may have just started out with a small biz idea – say just one shop or trade or craft, and then grown big organically over many years. I had read Made in America by Sam Walton of Walmart. Don’t remember right now, but his case could be one in that category. It could have been that only later did he think of huge growth as a company, not at the start or even for a few years.A good read, BTW, and so is Made in Japan by Akio Morita of Sony.Those were their respective autobiographies.

          1. srisays

            Yes you are right. I am halfway through Made in America, he speaks about relentless experimentation and focus on a single idea.

      3. kenberger

        As a sometimes composer, this is the ONLY way I knew how to write music, back when I used to do that.Which is why my tunes came out so infrequently.

      4. DJL

        You woke up with this in your head and I woke up with 10 chardonnays from the MIT Museum. That’s what makes you “that guy.” (Good seeing you and everyone.)

    2. PhilipSugar

      Better said than I could have.

  5. sigmaalgebra

    Well, yes, but whatever is done really does have to please enough users/customers right at the first to get the viral, etc., growth going. So in some sense there needs to be some content, some “beef” in the sandwich.I don’t know the early days of Google and Facebook well, but IIRC before Google was Alta Vista and before Facebook was MySpace. Now both the two earlier ones are gone or nearly so. Google and Facebook, right from the beginning, were enough better to start growing like a weed in wet ground in May.There are other examples of a not so good but early effort being totally wiped off the map by a later startup. E.g., Microsoft did not have the first PC operating system. Microsoft’s Windows was not the first PC operating system with a graphical user interface (GUI); IBM’s OS/2 was earlier. OS/2 flopped. Maybe OS/2 is in part an example of your idea: It long really was much better technically than anything Microsoft had before Windows NT or Windows 2000. But a huge problem of OS/2 is that the hardware it needed was too expensive for too many users and developers using Microsoft.ForIn startups, you get to show your stuff every day, all the time.Here I’m concerned about the standard “You only get one chance to make a good first impression.”.Well, Alta Vista and MySpace got shown everyday and got soundly beaten by startups that were better.For my startup, it’s all just dirt simple except for some crucial core math the users will never be aware of. The math is easy for me and has been in solid code for a long time. For the rest, that’s essentially just a Web site. It’s my first, and I kept is simple, dirt simple: Just HTML with only a little CSS. Microsoft’s ASP.NET has written a little JavaScript for me, but I’ve written not a single line of it. Maybe for running ads I will have to use/write some JS.There is no use of cookies, and users do not log in or have passwords. There are only two main screens, and they are each exactly 800 pixels wide, no use of div elements, and with layout precise to the pixel via HTML tables. The CSS usage is dirt simple. Except for the Microsoft JS file, a page sends as just one HTML file and one GIF file for a logo. It all sends for just 400,000 bits which means really fast page loading for nearly everyone.The pages are all simple and really easy to read and understand; fonts are large; contrast is high. There are some simple links as single words in English, e.g., “Help”, and some simple push buttons, but there are no pull-downs, pop-ups, roll-overs, overlays, or icons. In case a user has a small screen, there are scroll bars, and a user could get by with a screen as narrow as 300 pixels or have their Web browser magnify the screen by 2X or 3X and use 600 or 900 pixels.The pages are interactive but have only some simple one-line text boxes for input. The pages and interactions are so simple a child with no knowledge of English should be able to figure out the English version in about 15 minutes on their own or five minutes with an example from a friend.All a user needs is a Web browser up to date as of 10-15 years ago! We’re talking just dirt simple! The same Web pages should look good and be easy to use on anything from a smartphone to a workstation.Oh, right, every Web site needs a session state server. I used what Microsoft’s ASP.NET offered during development but wanted something better for production. So, I took out a little time and wrote a session state server. It has two instances of the standard .NET collection class, some object instance de/serialization, and some TCP/IP sockets. So far queuing is FIFO as provided by TCP/IP which, IIRC, can handle queues at least 100 deep which should be enough for a while. Writing that was easier than even reading the documentation for Redis.Starting with the session state server, ripping out the collection class code and adding a file write statement yields a site log server, again one I like better than what Microsoft’s ASP.NET has.There’s some light usage of SQL Server with a simple schema — no foreign keys, no joins, just a few simple tables.The key to the content being good and pleasing the users is the core math. That part was easy for me, and all the rest was quite simple software development, also easy for me. But there were unexpected problems, at this point overcome.So, the Web site should get a 10 for execution of the basic work because it’s all so simple it’s easy to have it rock solid and basically flawless. I mean, simple HTML table elements do work as promised, right? The code is 100,000 lines of typing and always so simple it was always routine and low risk.The main question is, will 2 billion people love it?So, I agree on the importance of “simple”. But the first version still has to be seen by the users as good, has to be good enough to make a good first impression and start viral growth.

  6. Tom Labus

    The swing of a consistent 300 hitter is usually simple without a lot of extra moves and gets great results. Great post, Fred.

    1. Mac

      I’m thinking Hank Aaron and Willie Mays. You?

      1. Tom Labus

        They would be great choices. Don Mattingly before his back went too.

        1. Mac

          Yes. Mattingly. Maybe something will come of the talk of him possibly returning to the Yankees.

    2. JamesHRH

      And they tend to be looking for something to hit, while fighting off pitches they can’t hit so they can get back to their focus.

  7. Mac

    I tried that approach once: reverse triple somersault with two twist in pike. Even threw in a full gainer to boot. Unfortunately, I misjudged the depth of the water.It would have helped had I had an MVP, enough runway, some founder/market fit and a solid team to pull me out in time. We live and learn.

    1. fredwilson


    2. Matt A. Myers

      As you said, having enough runway is generally the issue – that gives you time for MVP, market fit, build solid team, etc – the more runway, the more evolved the MVP, team, and understanding/vision you can create – and the longer the runway of aligning future investment (if revenues don’t cover growth).If you look at the biased view, the perspective, of a VC who needs to hear a lot of pitches – with hopefully other proof points, proof of concept etc – they of course need to hear the leading metric, they however don’t really have time to learn and understand a greater ecosystem, they may be able to envision what extensions to a business model will look like, however the founder perhaps has thought much more thoroughly through this.The structure of VC and the limited money they’ll give for a new venture, especially someone not considered inside the box of a serial entrepreneur, won’t allow for investment to allow for a runway that the larger holistic ecosystems require – which is why they seem to need their founder’s own capital as fuel to start.

      1. Mac

        I see I failed to mention I was blindfolded as well. As we’ve all learned, being able to quickly and consistently execute is key. Even if it’s, as Fred said, executing a few simple things perfectly a thousand times.

        1. Matt A. Myers

          Consistency comes with protocols being developed and followed, which comes with trusting people which requires learning from mistakes as systems and protocols are developed. This will lead to speed, of course. It’s learning what you need to execute perfectly for your circumstance that matters, I suppose that’s where proof of concept comes into play.

          1. Mac

            I couldn’t agree with you more. I’ve found ‘proof of concept’ a vital ingredient for long-term survival.

    3. Donna Brewington White

      This comment grabbed me, Mac.Was depth of water referring to something specific, like wrong target market. Or am I reading too much into it?

      1. Mac

        As usual, Donna, you’re spot on. I thought I’d executed the perfect leap into uncharted territory. Circa1996. However, my co-founder and I hit the bottom a wee bit hard. With no internet knowledge, no early-stage fundraising experience, no experience or track record in this space, no MVP, no traction and no team, we set out to raise an A Round of $1.3 million. Even a considerable amount of due diligence in our target market, along with strong relationships with industry leading outsource partners, wasn’t enough to overcome our ‘rookieness’. But, we looked good doing it. Maybe that counted for something.Hope you and all yours are well.

  8. Vendita Auto

    Advancing technology often makes the high degree of complexity necessary

  9. jason wright

    “a journey of a thousand miles starts with a first step.”who said that?

    1. Twain Twain

      Lao Tzu.

      1. jason wright

        so Mao was quoting Lao, …who was channelling Kao? 🙂

        1. Twain Twain


  10. JLM

    .Well played, the corollary of Occam’s Razor.The thing is you can always add complexity when you master simplicity.JLMwww.themusingsofthebigredca…

  11. jstylman

    In my office, we have a signed that reads: “Do Less Better” – it’s an aspiration that our instincts tend to violate more often than we care to admit. Along with posts like this, that visual is a great reminder to keep us in line.

    1. fredwilson


  12. Twain Twain

    I belong to the Da Vinci and Steve Jobs schools re. simplicity and complexity. In AI, there are lots of moving parts and there can be quite complex maths going on. Yet the consumer / user doesn’t need to see this.They just need to do one small quick action (e.g. click a button, say a word, type something into a box) and the AI engine is supposed to work things out for them. https://uploads.disquscdn.chttps://uploads.disquscdn.c…Unfortunately, the AI clearly can’t do NLU:*…So as a founder you have a choice: keep things simple and easy for yourself by building on top of the legacy frameworks — since everyone else is and that’s what raises interest and capital from investors (@fredwilson:disqus), even when it’s clear those frameworks are wrong and systematically discriminating against people.Or you do a Jobs and dig deeper to arrive at a better solution for everyone.

      1. Twain Twain

        Thanks, Arnold. The marketing is easy. The math logic for the AI is not. If it was simple and easy, Google & FB could solve it and they can’t:* https://developers.googlebl…*…*…Maybe we just forget about all of it and carry on making biased AI because that’s easier.And we’ll just spin the systematic biases against people into positives.

        1. awaldstein

          If marketing was easy everyone would be doing it well.Hardly the reality.

          1. Twain Twain

            If engineers can’t solve the math logic, there’s nothing to market. It’s like if Elon’s team can’t build the rocket, there’s no launching a Tesla into space for marketing wit.That’s the situation AI is in.

          2. Twain Twain

            Marketer: “We connect people.”The engineer …https://uploads.disquscdn.chttps://uploads.disquscdn.chttps://uploads.disquscdn.c…That’s just for the relatively “simple” problem of social networks.Natural Language AI and de-biasing the data (so that it doesn’t discriminate against people) is factorially harder as a problem.We may have to do those triple somersaults with a double pike out of necessity rather than for fun. There may not even be a theoretical paper showing how it might be done —much less a YouTube video and coaches who’ve done it before that the founder can ask and learn from.Maybe we just say, “The discrimination that gets propagated by the machines isn’t our problem. Let’s not do triple somersaults with a double pike. Let’s not try to move innovation forward. Let’s just market the heck out of existing straight dives because that’s all that the AI can do.”

    1. Matt A. Myers

      It’s the process of brainstorming – and then pruning. The more connections you can make, understanding nuances of what’s repeatedly shown as important, the more simple, logical a solution will seem.

    2. PhilipSugar

      Love this post.

      1. Twain Twain

        Thanks, Phil. Of course Fred’s point about KISS (Keep It Simple, Stupid) is right and we need to be vigilant about that. AVC regulars know how blockchain and crypto found it difficult to “cross the chasm” because they seem too complex and suffer from poor marketing. So l@SixgillBlog:disqus ‘s point about there being an art and talent to marketing is understandable.In AI, there was an assumption of starting with simple binaries and cellular automata to build up machine intelligence into human-equivalent intelligence. https://uploads.disquscdn.c…Binary was invented in 1701. Cellular automata in 1940s. By 2018, we have machine vision. https://uploads.disquscdn.c…Unfortunately, there are flaws in building up from those simple assumptions:*…Likewise, in Natural Language, the assumption was made we teach the machines simple words and build up more complex structures to enable them to parse sentences. https://uploads.disquscdn.c…However, those centuries of linguistic theory since Humboldt in 1820s and Chomsky since 1950s has resulted in:Google’s Fighting Hate And Trolls With A Dangerously Mindless AI*…This is an industry-wide problem. That’s why none of the big techcos can deal with the hate speech, trolls, “fake news” problems.Sometimes, building up from simple doesn’t get us to the right forms of complex.Maybe we need to dive into the complex to find the new simple basis. That’s what I think Steve Jobs meant.

  13. JLM

    .Of course, everyone knows that would have been a: “5233B”Every dive is designated by a numerical system.The “5” means a back dive. The “2” means back takeoff. The “3” means the number of somersaults. The second “3” means the number of twists. The “B” means pike position.I simplified the explanation because the somersaults and twists are actually counted in “halves.” Simplified.…In the future, you can just whisper: “5233B” Again, simplicity.JLMwww.themusingsofthebigredca…

    1. JamesHRH

      Who knew there was such a thing as inside diving?

      1. JLM

        .Who knew?The Shadow knew. The Shadow knows everything.JLMwww.themusingsofthebigredca…

  14. Frank W. Miller

    Its all about revenue. How does this filter back? The engineering staff should be making conservative, i.e. low-risk choices, when design decisions are made. Sometimes thats not possible, e.g. the high-risk technology is the point of the startup, but most of the time it is. If you want to try out some bleeding edge thing, do it in academia. That environment is specifically designed to try way out stuff.

    1. Matt A. Myers

      A lot of ideas get developed in academia like Google’s initial PageRank algorithm as an easy example; in hindsight it’s pretty amazing that something like the PageRank algorithm hadn’t existed before.Likewise, the structure of most VCs thinking and processes are happy to see revenue models (existing enough for proof of concept or through a VC’s experience with similar business models), however as soon as ideas are outside their box – either in terms of context or size/scope/breadth, their attention (which is required for learning and building trust) will be overwhelmed, leading it to dissipate. Big big ideas will therefore be outside their scope of risk-tolerance, even though big big ideas will scale out similarly to smaller ideas. Smaller ideas on the surface seem more manageable, requiring less investment in terms of money and perhaps time allocated, however in reality big big ideas require a relatively equal amount of money and time investment… it’s just likely they’ll be used to certain levels of proof of concept that a smaller idea can perhaps more easily attain, than if building for a more holistic vision – which of course will benefit from greater economies of scale as well.Anyway, the internet is broken because of this, society as well. Capitalism has failed to bring resources to those who have the minds for big big ideas, save a small few such as Elon Musk who’s path put them into a position of being able to support the launch of their own big big ideas; who do we go to for our own big big ideas when we only say have a maximum of $1MM to put into a big big idea that likely requires 10x that for a fair start, in order to make sure all the cogs in the machine are able to be put in place when they’re needed?

      1. Frank W. Miller

        I’m actually talking about it from the entrepreneurs pov. If more company owners focused on low risk execution and getting to revenue quickly, they wouldn’t have to take that crazy expensive VC money…

  15. JaredMermey

    Optimize the input to output ratio.

  16. Vitomir Jevremovic

    A long time ago, I got the best life advice I could have hoped for, but haven’t grasped it until years later: “You are always talking about millions, while tens of thousands are passing right by you”

  17. Donna Brewington White

    In business. In life.For so long I thought complexity was a virtue. Or at least made things more interesting.Age cures that, some. Or maybe it is the wisdom and shorter runway that come with age.

  18. Pete Griffiths

    Not only that… You can claim you had the whole thing planned from the start, look like a genius and mislead countless enthusiastic young founders.

  19. cavepainting

    True and profound wisdom in these words.Doing hard things appeals to the ego. It is like kids who try the hard flip to impress his friends and parents even though he knows he is likely to mess up. It feels more powerful to want to try something expansive.But it is not pragmatic and almost always ill advised.If you are deeply in touch with the mission and the problem to be solved, you want to sequence it in a way that solves for progress and eventual success. Do not solve it for the ego, how impressive it looks, or any kind of idealistic milestone that is not practical.Big things get built brick-by-brick-by-brick with sweat, humility, and focus.

  20. FinTech Connector

    Simply said but for many so hard to execute. I come across many fintech and blockchain startups that drown in complexity and never truly articulate the real problem they are trying to solve. Fortunately, for them in this environment right now the more complex they sound the more money they get – go figure!!!

  21. howardlindzon

    indeed. You win from polishing your game quietly in a corner of the world with a great angle of attack

  22. Diamondblade

    thats what ive been told by some succesful startups owners: 10 years is needed to go throuhgh all ups and downs ant then you become a master.

  23. Kirsten Lambertsen

    Great metaphor, which I’ll steal for conversations with clients. I deal with explaining this to clients almost every day.

  24. Jeremy Robinson

    Simple repeatable steps. That was the coaching philosophy of basketball legend John Wooden. A mentee of his whom I coached reminded me of it. I try to adhere to this method as a leadership coach. Focus. Reduce information over load. Simple repeatable steps.

  25. ShanaC

    I think I needed to hear this

  26. jason wright

    forward rolling rolling forward.

  27. Guy Lepage

    Great post.. “…execute, nail it, then…” I can’t emphasis that “NAIL IT” part anymore.. I feel a lot of startups miss the nailing portion. They feel a few individuals liking something is “nailing it.” I’ve witnessed more and more engineers releasing a product or feature and then just skipping over the nailing it segment.Products and features must be Developed (to make them work), Designed (to make people want it), Distributed (to make it known and available to the masses).

  28. JGreenberger

    @fredwilson:disqus love this post and I actually think a lot of these principals may be applied to management as well – just took a stab at writing down my thoughts:…Would be curious to see what you think in case you have a second!

  29. Richard Garand

    Here is exactly why so few want to do it: “But you will have done it by doing the easier things perfectly thousands of times instead of the hard thing just once.”It’s a lot easier to take one big shot and hope it works out. It certainly sounds easier than taking thousands of actions. It’s a lot more exciting because you get to be the daring person taking a huge risk (for the benefit of everyone else) who will be glorified endlessly if it works.But free markets tend to be the ultimate test of whether you will make the hard choices and there’s no faking it. So ironically there is less competition when you actually choose the right way.