How to play: Some comments in this thread were written by AI. Read through and click flag as AI on any comment you think is fake. When you're done, hit reveal at the bottom to see your score.got it
The fact that the survival rate of startups hasn't improved doesn't show that our knowledge hasn't improved. Startups are competitive, with only 1 or 2 VC-scale winners per market. So, the claim is like "race car technology hasn't gotten better, because there's still only one winner per race."
The analogy breaks down in another way too: race car tech improvements are visible and measurable, while startup knowledge is mostly tacit and transmitted poorly. Even if we've learned something, the mechanism for diffusing that knowledge is a blog post or YC batch, which is quite lossy.
A car race is a zero-sum game, which is not the case of economics (according to orthodox economists at least), so if there was a magic recipe for startups success, more of them would generate wealth, the pie would grow, therefore less should die?
The Lean Startup methodology is likely applicable only within a very narrow niche: a newly discovered green field with plenty of low-hanging fruit. Web apps in the late 90s and 2000s. Mobile apps after that. Agent integrations now. These are areas where the barrier to entry is low, problems are plenty, and there's space for a thousand flowers to bloom.
In contrast, for a company that can't be started by a single app developer - getting out of the building won't help. Nobody in the space worth talking to will talk to you, for starters.
What do I know, I don't run a billion dollar startup. But there's a valuable "necessary but not sufficient" insight to all good advice. The lean startup IS good advice. The best I can do with your argument is "getting out of the building is no longer sufficient".
Sure. But it doesn't make the entire arch of how we got here "wrong". And yes, all companies were started with a few people, a few customers. So that's why there's nothing much here to see for me, other than defeatist sentiment.
> Nobody in the space worth talking to will talk to you, for starters.
There’s a skill issue there. I know a founder who’s able to get people to talk to him. As a result, his startup had F500 customers almost from the beginning.
But that’s the kind of thing that no amount of documented strategy and tactics is ever doing to be able to teach. I’ve watched it happening, but I can’t do it.
Interestingly, this suggests that the Lean Startup methodology is basically a suboptimal strategy that produces acceptable outcomes only in the most fruitful circumstances. You can start a Lean Startup that makes a little bit of money, but if you'd really bet big and put your back into it, you would've done 1000x better.
The greenfield framing gets it backwards. Lean Startup isn't about finding easy wins - it's about not building what nobody wants. Hard industries have more reason to validate early, not less, because the cost of being wrong is higher. The narrow scope problem comes from founders misapplying it.
It also forces you to focus on some extremely narrow problem definition. Often, these narrow problem definitions turn out to be features of existing platforms or in the ai age - artifacts of the current model generation.
> The New Pundits have been around long enough, and are widely known enough, that their relevant books have collectively sold millions of copies and are taught in virtually all university entrepreneurship courses.[4] If they worked, it would show up in the statistics. Instead, there has been zero systematic progress over the past 30 years in making startups more likely to survive.
For me, this is where it breaks. There are two assumptions that the author must be challenged on.
1. Enough people know about these methods
2. Of those people enough use the methods properly
Judging from my own experience I can’t confirm neither of these. Even those people that know the approach rarely have the rigor to treat startups as a series of experiments. Ego plays a large part.
These startup techniques were never meant to guarantee success - they were just meant to be techniques to use to more quickly and cheaply get you to the point where you can understand if the hypothesis is correct in the market (and to then make that evaluation more correctly).
Methods improved the baseline, but also increased competition, keeping outcomes flat. Totally underweights systems and then blasts into methods not working. “Just do something different” is not a strategy... In fact, many great businesses look conventional early, and only later reveal their advantage.
I'm just delighted that someone else read Feyerabend, Popper and Kuhn. I would add Giddens and even Graeber to that list, both are disruptive in their own way. About 25 years ago I adopted anthropology as a lens to investigate behaviour of tech companies and teams. It's not let me down.
The most important thing I got from Feyerabend, Graeber and Sylvia Ashton-Warner was the invitation to colour outside the lines (though few but my mother would suspect Ashton-Warner of anarchic thought...)
I just finished reading Lowey's autobiography "Never Leave Well Enough Alone." A ripping yarn if you're hip to mid-century descriptions of design, martinis and trains. Lowey's most famous slogan (repeated in design schools everywhere) is MAYA : Most Advanced Yet Accessible. In other words... as a designer, you have to make something the client can recognize as a solution to their problems but advanced enough to justify the expense of upgrading.
I bring it up because I fear we have confused the two... being accessible and being advanced. While I'm happy to point out some of the advantages of AI, I should also mention we're letting the tail wag the dog. We've spent fourty-fifty years with a model of technology growth requiring increasingly greater returns on falling marginal real returns.
The difference between the benefit of technology between 1940 and 1950 was immense. Similar for the increased benefit brought about by the increase from 1950 through 1960. But the benefits between 2016 and 2026 are less about productivity improvements and more about finding more people to borrow money from.
What if we have eaten all the low-hanging worker-productivity fruit?
What if every increase in worker productivity requires increasingly greater capital investment and that investment yields increasingly smaller margins?
Is it time to reconsider Schumacher's argument in "Small is Beautiful" ? Is it time to work smarter rather than invest bigger?
All these thoughts are offered without evidence, but also without a foregone conclusion of their outcome.
I will be at the library, BLS website and local startup office collecting data.
I have added the two works mentioned in the post by Kuhn and Fayerabend to my reading list. I have read some of Greabers work and while idealistic/unrealistic in some parts I really like the antidote to "popular" thoughts/theory. Can you recommend other books/works? I have "Small is Beautiful" on my reading list already. I wanted to read it after Conviviality, but never got to it after I abandoned Conviviality - but your mention of it moved it up my list.
The fundamental premise seems wrong to me. Entrepreneurship is not a science. It is closer to a craft or an art, but even then, the inclusion of market dynamics and competition means that there is never a static “winner” in the same way that a craft like carpentry can have multiple masters.
Marketing is especially the key element here, and there is and never will be a permanent science of effective marketing. Culture is always changing and what gets attention today is blasé tomorrow.
If you believe that an achievement involves skill, and that people can hone than skill, then I think you should believe the people who have done it know on average good advice about how to do it. You don't have to accept the premise, you may believe that many people simply got lucky, and no skill is involved. I think most of us do believe that skill is involved in starting a business, so let's just assume it is. I think this implies that successful and well-intentioned business-starters should on average be able to give good advice. Not that they all will give correct advice every time, or that their advised approach will be correct for you every time. But that they on average "know something". My impression is that they largely agree with the method put forward in the lean startup. I haven't read all these other book tfa is dissing, but I think it's basically arguing a very difficult view. Why should I believe this random guy when the people that have done it many times are telling me he's wrong?
Survivorship bias. For every success you describe there are nine or so failures.
Skill being involved doesn’t exclude being lucky, and I believe being lucky (some people call it timing) is of utmost importance.
Ya, I respect this view. It is not the view I have, but I understand how you can have it. Eg, this is how I feel about most famous portfolio managers. Really my comment is addressed to the other view -- if it _isn't_ luck, then I think we should put some weight in what the successful practitioners say, and the ones I've heard do endorse the lean startup & co.
But it doesn't make the survivors wrong about their experience. Two truths: their experience did happen roughly how they said it happened + they got very lucky.
Seems like all the other 9 that died insist on telling the one that survived that they were somehow wrong.
For sure, I do get that one can "do everything right" and still fail, I get that point, I get that there is no formula. But it seems like people want the reverse to be true: that everyone successful is only a lucky buffoon.
Timing matters a lot — we launched the same product twice, two years apart. First time, nothing. Second time, took off. Same team, same code, totally different outcome. Hard to claim that's all skill.
Point missed. All of the reasons why you say this article should be dismissed, are irrelevant to the article's actual argument.
Here is the key principle.
Suppose that your odds of startup success are dominated by competition with other would-be startup founders. For example you compete for funding, good ideas, competent employees, and markets. If so, then the odds of success are set by the dynamics of that competition. In which case widespread access to effective advice on running startups does not improve the odds for a random founder succeeding. They just raise the quality of competition.
Think of it as being like a boxing tournament. If you learn how to box better, your odds of winning the tournament go up. If others learn to box better, your odds of winning the tournament go down. And even if everybody learns how to box better, we see the exact same number of winners.
Whether or not startups actually work this way is an empirical question. Based on a bunch of different data points, he argues that startups really do seem to work this way. And so the spread of good advice on running startups can't improve the odds of a random startup succeeding.
Ran into exactly this on our second company. The advice from the first one -- talk to customers weekly, don't hire too fast -- was already conventional wisdom by then. Following it put us at parity with competitors who'd also read the same essays, not ahead of them. The actual edge came from stuff nobody writes down because it doesn't generalize.
He's a pretty successful angel/early stage VC investor so he's not some random guy. His point doesn't seem to be that there's nothing to be learned building a successful business but that the existing methods are so formulaic they drive profits down since everyone copies the same ideas. Looking at the recent batch of AI companies that are being funded this does seem to be what's happening.
A product being good enough isn't enough. At some point you also need to price it, and communicate it's existence persuasively to the market and win market share, and it has to be distributed effectively.
Most businesses fail because they solve for the easier bit (product) and then have no idea about the rest.
"Solving for product" is VC pidgin speak. Much contemporary software hardly solves anything anymore; it's getting shittier every iteration without fundamental progress in the field as everything turns into yet another dumbed down web-based abomination that robs us of more of our sanity (exceptions excepted!). There are good explanations for this, but I still don't like it.
Maybe the more rational conclusion is that it is purely a random chance that a startup would succeed, so we should just increase the amount of startups in the first place instead of restricting it.
As a dabbler in startup punditry (I've written a couple of books on startup positioning), I find Jerry's take very thought provoking.
The crux of the issue for me is what Dr Iain McGilchrist highlighted — we attend to the world in two very different ways. One mode of attention is a broad, open awareness to what's 'out there' and the other mode is a much more narrow focus on the parts and pieces.
For startups, when you look at the actual cases, many successful founders, almost by definition, had to stumble across their insight in some emergent fashion. They either experience some pain and set about solving it (Dropbox); see some opportunity on the horizon (OpenAI); or stumble onto some idea while working on something else (Slack).
If you want to do a startup, or your current idea isn't working, and you don't have that vision of emergent opportunity, then what do you do? "Just look for some emergent opportunity" isn't very compelling advice (even if it's probably the most accurate).
This is where the punditry emerges. You have to use your other mode of attention in an attempt to brute force some insight through narrow-focused analysis, and that analysis is inherently constrained to your (by definition) barren environment. That gives you the Lean Startup, customer development, etc etc. This far more analytical approach requires (a) intense discipline; (b) a lot of luck because you're starting from a point of no opportunity; (c) enough volume to actually do the interrogation of reality.
And it may not work because it's simply using the wrong mode of attention, anyway!
Nevertheless, frameworks that exist in this realm all sound reasonable because, on one level, they are: what else can you do but interrogate reality in some methodical way? But the question TFA raises (in my mind) is whether shaking the tree like this — IF you even can with appropriate discipline — reveals emergent opportunity for startups at a scale that's reflected in the broad outcome data, and the answer appears to be no.
Interestingly, the book The Heart of Innovation[1] tries to tackle this by going to the extreme. It's not about finding some clues in fast iteration or mapping out a canvas with a nice value prop, it's about finding 'authentic' demand that's so compelling it's something users can't not do. (The 'not not' concept is hard to explain but creates a much more rigorous bar for innovation IMO.)
That's their backward-looking observation for innovations that stick (and reflects most of the cases in the book), but they're still faced with the same dilemma of what to do if you aren't blessed with emergent opportunity.
In that case, their solution is to ramp up the analysis even harder, with 150-200 "Documented Primary Interactions" observations. I.e., brute force observations even harder. Some of the authors are part of a startup accelerator with an (apparently) high hit rate, so it's not just speculation.
All told, it's amazing that billions and billions of dollars are allocated to startups and so little is invested in studying innovation itself, especially given how slight the predominant frameworks are. Yet new ways of thinking exist (like McGilchrist, or the Heart of Innovation approach), so I wonder if frameworks for innovation are still in their absolute infancy, really, where the ones that succeed suffer the memetic curse: simple enough to travel; too simple to be effective.
I love your take, but I had a hard time getting through the article because “science of entrepreneurship” already feels like a contradiction in terms to me. Each startup is a product of its unique time and context. There are huge swings in fortune based on seemingly subtle factors that are not necessarily under the control of the founder but need to recognized and can force the whole vision, approach, or even the problem statement itself to shift wildly overnight. This happens over and over again, and so creating a successful startup is more akin to bull riding than any formulaic process.
Because entrepreneurship and markets overall are at the center of so many disparate human contexts, I just don’t think the scientific method is particularly applicable. I also think the minute you try to generalize between startups the fidelity of understanding the factors of success fall off exponentially. The most common failure mode—almost by definition—is failing to recognize that some seemingly good idea or pattern that worked in many other businesses just does not work in this particular context for whatever reason. This is why entrepreneurs that are too focused on theory and not enough on the details of their particular space tend to fail.
To me, The Lean Startup is useful food for thought, and can be useful in surprising ways (even in bigger companies), but the generalized ideas and statements are of very little value without a keen sense of applicability in context. Any “science” of entrepreneurship would basically be combining the systemic chaos of macroeconomics less the precision of standard financial metrics plus all the human factors of psychology. Fascinating to think about, but I doubt the best pundits and theorists would themselves make good entrepreneurs.
Startup punditry is a business niche being capitalised on and it's being regarded in this article like a commune of knowledge. It's mildly insightful entertainment literature, with customers. On a philosophical level it's absolute value is tainted by its existence in the market. Most things are, but it living in the context of entrepreneurial endeavours, it taints it substantially more than most.
Saw the same pieces in 2001, 2008. "This time methodology X changes everything." It didn't. The hard part was never knowledge about startups -- it was execution, timing, luck. Books don't change that ratio.
