Lean Startups suck. Here are 10 reasons why…

Eric Ries’ 2011 book “The Lean Startup: How Today’s Entrepreneurs Use Continuous Innovation to Create Radically Successful Businesses” has garnered lots of attention, and indeed plenty of favourable reviews. He’s a thoughtful guy, who has been gradually building up interest in his “Lean Startup” ideas over the last few years: if you haven’t really heard of it, it’s basically an abstract conceptual toolkit aimed at very young companies that offers plenty of sensible-sounding advice, with very little to obviously disagree with.

But even so, I think that Lean Startups still suck. And here are ten big reasons why:-

(1) The “Lean Startup” approach is not a science. In fact, even though much of it is couched in terms such as “empirical” and “hypothesis”, it’s not even properly scientific. What it offers is a load of consulting-style models mixed in with contemporary truisms about business presented as hypotheses, which Ries hopes you will then go out and test for yourself. Why? Because despite all his evangelism, so few businesses have actually tried being lean startups that he hasn’t yet got enough data points to construct convincing arguments for his book. Which is why he so wants you to go test his ideas for him with your money and your business on your customers. If his ideas were scientific, they would have been tested already: but because they’re not, they’re merely hypothetical, which is a different thing altogether.

(2) “Lean Startup” is not a Scientific Management theory. I’ve seen Eric Ries talk, and he likes to claim Scientific Management as a (retrospectively adopted?) parent of Lean Startups, but in practice that’s simply not so. True, he likes metrics (and Frederick Taylor’s ghost would surely approve of that): but metrics and dashboards have been in high management theory vogue for well over a decade, so they are merely things that Ries has appropriated rather than devised in any useful way.

(3) “Lean Startup” is not a Quality System for R&D. The core of Ries’s claims is that his Lean Startup ideas provide a more productive way of getting to “product/market fit”. However, in its current state it is not even remotely close to an ISO-9001 approach for systematizing R&D: as a general rule, I’d say quality systems are a poor fit for R&D companies. Rather, the key test Ries applies to everything is relentlessly Darwinian: is your product a good market fit yet? (If not, iterate and try again). But as far as I know, there is as yet no substantive academic research on whether relying so utterly on customer feedback is in fact a good way to design new things: it could be argued just as strongly that it’s merely an effective way of making incremental design changes.

(4) Lean Startups are unfundable by angels. I estimate that 95%+ of current business angels (oh, and 99%+ of VCs, too) would not currently invest in any lean startup. This is because the Lean Startup toolkit has a single operating model: a Zen state of sustained-yet-aggressive ignorance while you intuitively develop some kind of tentative product-market fit by repeatedly bouncing things off enthusiastic early customers. Even though this may be in some ways inevitable if you’re trying to build a startup using your own savings, it is anathema to almost everyone looking to back a startup. I’ve talked with 130+ business angels, and almost without exception I would say that they want to put money into small, highly productive teams who have the self-belief that they already know what they are doing, so that those people can get on with executing their plans and making them work. Angels don’t want to fund your industrial education, they want to fund your market-focused business.

(5) Lean Startups are not lean. That is, as far as I can tell the ‘lean’ part of the title was merely appropriated from “Lean Manufacturing”, which - I would argue - bears almost no relationship to Eric Ries’s customer development ideas (which were instead built to a very large degree on Steve Blank’s Customer Development ideas). In fact, Ries’s idea’s correct title should probably be “The Learn Startup” because it is 100x more to do with learning than with reducing process waste and quality management: it’s certainly not to do with doing things cheaply or even fast, because educating yourself into a constantly changing market via progressive iteration can take an entire lifetime, and can happily absorb every penny you care to throw at it.

(6) Lean Startups makes little sense outside Web 2.0 startups. If you are considering setting out as a “lean startup” manufacturer, I’d urge you to think again. Iterate with your customers all you like, but you still need to pony up heavy-duty capital when setting up a manufacturing process, and the Lean Startup consulting toolkit is of little use when funding or developing this kind of thing. Regardless, you’d probably be better off reading up on mechatronics development, because this covers an awful lot of the same area (did I mention that it’s not new?) without telling you that you need to iterate in order to build worthwhile things.

(7) “Lean Startup” ideas have not yet been tested. Look, anyone – and I do mean anyone – can assemble a set of contemporary-sounding statements about how best to run a business, and then claim that the multiple similarities between those statements and carefully-selected details of the runaway success of { Google | Facebook | whatever } form cast-iron proof that your overall theory works universally. Eric Ries knows (and indeed points out) that this does not amount to any kind of real wisdom: however, because he has as yet amassed no real data, I think this is exactly what his ‘Methodology’ reduces to.

(8) Lean lacks cojones. Or “bravery”, if you wish. By delegating design decisions purely to the result of customer iterations, it encourages startups to produce incremental me-too wannabe products that appeal only superficially to customers, and to make design decisions based more on incrementalism than disruption. As such, it is a way of developing things that is intrinsically anti-engineering. Yet I would argue that to build company value and have lasting societal impact, startups need to be (quite literally) revolting.

(9) Lean is not a movement. It is not even a popular revolution. It’s one guy touring the world telling developers telling them how smart they are, and selling them an unsustainable dream – that it is in their power to build their own company cheaply and effectively by iteratively running prototype product versions past early adopter customers. It’s a tempting notion – after all, it’s the same basic dream that fills hot desk seats in Old Street and all the other tech hotspots worldwide. But ‘ought’ doesn’t make it ‘true’, no matter how many times you say it to audiences who would like it to be true.

(10) It would be nice if Lean was truly as good as Ries claims. But it’s not. In some idealized parallel universe, it might well possibly be true. But in this particular universe, it is at best no more than a technical hack for trying to avoid really big customer-facing design mistakes. So, unless you (and your optimistic co-founders) can entirely self-fund your killer startup through a (probably very prolonged) lean development phase, I would strongly advise taking your Lean Startup noodle soup with a large pinch of salt.