It’s got to be the single-most asked question in the software business world. Beginning startups – both rolling in VC hay and bootstrapping – ask it. Existing software companies, as they get whacked around by a changing market (Mobile, where’d that come from?) and changing user expectations (You mean, I don’t just run in my browser?) ask it. The polite form of the question is, “How do I/we define what product to create?“. What they’re really saying is, how the hell do we invent (or re-invent) a software product that will sell like mad before we go broke like in out of business?
The traditional way to cope with this is,
Except for the other 9 out of 10 startups whose software arrives to a vast collective yawn and are dead meat in 3 months.
A lot of smart people who’ve been through the above have wondered if there’s a better way to break into a market than running blind at it headfirst. And smart people like multiple startup founder/academician Steve Blank, Marc Andreessen (yes, that Marc Andreessen) and a serial startup founder by the name of Eric Ries have come up with what might be an answer – and might work for you. Apply the principles of lean manufacturing to building startups, lean startups.
The core idea, in my opinion, of lean startups is getting to the right intersection of Product with Market. That’s the detonator cap to making it possible for the company to explode into profitability. Unless you obsessively focus on getting that product/market fit first, having great programmers, tons of VC money, storied executives etc. etc. is about as useful as pounding on a brick of Semtex with a mallet: nothing happens. (Note: do not try that at home. Or anywhere else.)
Getting to product/market fit is about getting the data every step of the way of what the market will think of your productized idea. Call it product development based on market research, validation, feedback – on steriods. Your progress until you find product market fit is measured by units of validated learning about your customers, not milestones, integration tests or cases of Jolt consumed.
Here’s a quote from Eric Ries’ post for GigaOm that neatly sums it up:
As a consequence, this new startup is relentlessly metrics-driven. It tries out new ideas with a fraction of customers in order to prioritize using facts, not opinions. Its unit of progress is that of validated learning about its customers. Because this radical notion of progress is located firmly in the heads of its employees, and not in any artifacts they produce, the lean startup is employee-centric and knowledge-obsessed. It is a truly fun place to work.
This radical notion of creating software people actually, measureably want has a lot more to do with Internet’s ability to connect everyone, measure everything and do it for free than the industrial model of software development (typified by giant software projects that with predictable regularity crash and burn. )
I interviewed Eric Ries for my podcast recently – the show should be up before 2010. The thing I got most out of our talk – other than becoming a believer in Lean Startups – is that Lean Startups scale down as well as up. This is the good news microISVs and small startups have been casting about for: there is a way to find your way to the interestion of product and market.
Here’s a few good resources to actually learn what Lean Startups are all about.