Finally I got round to reading Eric Ries’ well known book - the Lean Startup. This book often creeps into conversations and conferences so I was keen to see what was in there rather than relying on other’s insights.
From a developer point of view, it’s a fairly business-ey read as you might expect but there’s some great stuff in there and lots (and lots!) of examples from the technology industry as well as more traditional industries. Here’s my main take homes from the book:
Build-Measure-Learn
In the highly uncertain world of a startup you might have a good idea or a product but how do you know anybody wants it? In short; you don’t. So what can you do? Ries suggests you make an assumption of what the customer might want. Then figure out how you’re going to measure it’s success. Build it, ship it and then review the metrics. Did it work? Didn’t it? Most importantly what did we learn? The important thing is not building a truly awesome polished product; it’s learning what the customer is actually after as soon as possible. Write throw-away experiments or establish a minimal viable product to provide enough of a basis to get learning.
Vanity Metrics
This term was new on me and it seems to have really stuck. These are literally everywhere. Ries believes that all too often people make assumptions based on vanity metrics. A good example might be using the total number of users as validation for a product’s improvement and increased popularity. Let’s say the development team released a new user registration page on some date and there’s a spike in the total number of users, that must be because the lovely new layout, right? Possibly. Or maybe marketing started a new advertising campaign on that date? There’s probably lots of reasons all of which can’t be fully determined. This doesn’t mean vanity metrics are useless though. Using A/B Testing you could compare your vanity metrics on the two versions of the product and see which one performs better. Perhaps releasing one change for a week to half of the users to see if they’re more likely to purchase something.
Innovate
Companies are often formed based on an innovation. However, they soon get bogged down with the operational and infrastructural efforts which go into maintaining and improving their established product. Ries understandably argues that innovation must continue or an organisation will face stagnation. How do you successfully create an innovative environment though? Ries suggests that “strong cross-functional teams” should be allowed to develop around an innovation. These teams should have all the resources required to see a product all the way through end-to-end. Then once the product moves from phase to phase there could be a hand off between teams allowing others to take on the further improvements and maintenance. Further to this, allowing the original members to stay with that product or move on to something else. Re-reading this it actually sounds quite similar to that described in the Valve handbook for new employees. Slightly less ‘Valve’ is Ries’ vision of having ‘Entrepreneurs’ within established organisations, presumably to come up with all the ideas. Not sure about this one.
Waste
Towards the end of the book there’s some great insight into the much wider problem of waste. Ries says “In every industry we see endless stories of failed launches, ill-conceived projects, and large-batch death spirals. I consider this misuse of people’s time a criminally negligent waste of human creativity and potential… What percentage of all that waste is preventable?”
Which of our efforts are value-creating and which are wasteful? Bare in mind that value is providing benefit to the customer. How often do those of us who are software developers actually benefit the customer? Personally, as a developer of APIs and not consumer facing products, not nearly enough. Our products are abstracted away from the customer so much so that it’s very easy to lose sight of their needs.
Finally, I’ll leave you with this great quote from Peter Drucker also taken from the book - “There is surely nothing quite so useless as doing with great efficiency what should not be done at all”.