Blue Flower

Lean Startup Theory

If there was ever a theory of organisational effectiveness that deserved the description 'buzzword', it's this one. Beloved of Silicon Valley and private company execs throughout the world, it refers to a particular way of creating and running new companies.  So what is 'lean' (as its rather fervent disciples like to call it)?

The classic example of LST always given is of someone in the US who thought that people might buy shoes online if given the chance.  With LST, you don't wait for the perfect megaproduct/project to come along, but get on with launching something small-scale (a 'Minimum Viable Product' in LST jargon) to test whether the idea you have has any chance of working.  The sooner your MVP is out there getting feedback, even if the evaluation processes are very rough and ready, the easier it will be to move up to the next level.

So the would-be shoe seller, rather than starting a company selling shoes online, building a big automated ordering system for it and so on, he simply took photos of shoes at his local shoe store instead.  He then put them on a simple website and, if anyone wanted any, went down and bought them from the store at full price.  When he was getting more orders than he could afford to subsidise the postage on, he knew the idea was a goer.  Unkind critics might think this is a long-winded way of describing the term 'pilot project', but LST enthusiasts insist it's more than that.

The concept of 'pivoting' is also particularly important to LST.  If the feedback on your MVP isn't good, you just change it to whatever it is that the feedback is saying would work.  The concept of 'validated learning' is also crucial - basically evaluating as frequently as possible (days/weeks rather than after months or years) and improving what you're doing after every evaluation.  According to the founder of LST, every bit of your organisation should also be treated like its own individual startup, with budgets divided accordingly.

So does it translate to a way of running social organisations?  The short answer is no.  Starting up a social project to help (say) mental health service users that you know is flawed raises moral objections in a way that a flawed shoe sales website does not.  Pivoting from providing support for (say) mental health service users to support for (say) mothers suffering from post-natal depression also affects people's lives a lot more than a private company flipping from selling shoes to (say) money off vouchers over the internet.  From an evaluation standpoint as well, validated learning sounds good, but in practice it's often indistinguishable from bad evaluation with a very limited sample of users - there's always a tradeoff between swiftness and quality when it comes to measuring impact.

With all these business guru type approaches too, the nice shiny theory is also not usually the most important part of the story.  In the case of the online shoe seller, it was good old fashioned nepotism that actually turned his business from a startup into a proper company, as when all else had failed he was able to borrow $150,000 from his friends and get his brother's father-in-law to introduce him to some venture capitalists he happened to know.

But forewarned is forearmed, and there are very few theories from which absolutely nothing can be learnt - so if you want to make up your own mind, this guide from the founder of LST is as good a starting point as any.