Startups: How to Iterate Faster
Release early, release often has become the mantra for startups. But how early is too early and how often is too often? Does speed really trump quality? Jeff Atwood, co-founder of Stack Overflow thinks so.
But is speed vs quality the right question to ask? High quality code is by definition more maintainable and more extensible; it is what you need to be able to iterate faster. When it comes to quality vs speed; is it 'either-or', or is it 'both-or-none'?
"Startups that succeed are those that manage to iterate enough times before running out of resources" — Eric Ries, the creator of Lean Startup methodology.
So how can startups go faster if they do not compromise on quality? What else can be compromised to gain a speed advantage? Paul Graham, co-founder of Y Combinator, provides one answer: "By 'release early' I don't mean you should release something full of bugs, but that you should release something minimal. Users hate bugs, but they don't seem to mind a minimal version 1, if there's more coming soon."
Initial product specifications come from founders' vision, not from a feature list from prospective customers, according to retired serial entrepreneur Steve Blank. The startup's first focus should be to test that vision and learn. Hence, their minimal version 1 should be only features that are core to that vision.
Early customers are those who would use a product in its initial stages given its' limitations and pain of adoption. They are the most important customers, because they will tell the startup whether there is a market for their idea or whether they need to go back and iterate. After all, startups don't fail because they don't have a product, they fail because they don't have customers and a market. "Greatest risk is not development of new product, but development of customers and markets" — Steve Blank
Startups can iterate faster by postponing anything that is not related to validating the vision. For example, Groupon started out as a WordPress blog with coupons available in PDF form. The first launch of the iPhone did not have Cut-And-Paste nor searchability of contacts. Integration with 3rd party and supporting multiple platforms are other examples of features that may not be necessary during vision validation. When the Chrome browser launched it did not support Mac nor Linux.
The above graph show the Technology Adoption Life Cycle Curve from Four Steps to Ephiphany. It is not until the vision is validated should the mainstream market and crossing the chasm be tackled.
Follow Technorati