What if I told you that your startup can avoid making a product, it just needs to suggest that it might make a product and see if anybody agrees to buy it? You might think your humble author to be mad. What if I told you that a former CTO of a startup that never managed an exit wrote that on a blog, and then again in a book? OK, now the idea is plausible!
That is not, of course, the message behind “The Lean Startup: How Today’s Entrepreneurs Use Continuous Innovation to Create Radically Successful Businesses”, but it is the part of the story that resonated the loudest. Eric Ries had spent a long time with his company, IMVU, building a 3D avatar experience for chat clients that it turned out people did not want to use. His insight was that the company could have discovered much earlier that people did not want to use it. They did not need to build it to discover that; they just needed to ask.
Coming first in blog form in 2008, and then book form in 2011, “The Lean Startup” distilled lessons from other authors, notably Steve Blank, into pithy, actionable advice. More serendipitously it landed at the perfect time for investors and technologists looking for advice on how to make technology companies on the cheap. The dot-com bubble’s burst was still in recent memory, with VC investment hitting its lowest ebb in 2003. Venture capital availability had turned down again in 2005, with investors preferring to support later-stage companies, and then 2007-2009 brought the Great Recession, with the property-led Global Financial Crisis at its peak in 2008. In short, there was not any money, so you may as well spend what little you had on a book that told you how to make do without any more.
The core message of the book is “pivot or persevere”: gather data that validates or invalidates your current course and pay attention to it. If customers are not excited by what you are promising, promise something else. If people react well to it, then carry on—but look for the next opportunity to test your assumptions.
In this, we see the seeds of some of the great technology companies of our time. For example, Justin.tv started as a YouTube-like video broadcast site, but pivoted to Twitch.tv in 2011 when they realised that videogame streaming was their most popular category. Burbn was a mobile check-in app created in late 2009, until the founders realised that their product was too similar to FourSquare, jettisoned all of the features except the popular photo-sharing, and chose a new name for the service they relaunched in 2010: Instagram.
The lean startup approach also brings many phrases that are in common startup (and other business) parlance today. A brief list:
- Continuous deployment: ship your experiments (and their outcomes) to production as quickly and as often as possible. Yes, we already had continuous delivery in the manifesto for agile software development, but this time we mean continuous and we mean delivery; we do not just mean running your tests when you make a commit.
- Minimum viable product: identify the smallest thing you can build that gives you high-value information on whether your approach is a good one, and build that and get it to customers.
- Split testing: now more commonly called A/B testing, this is the practice of randomly assigning people into different groups and giving them different experiences, comparing the outcomes on some important metric and searching for statistically significant differences.
With hindsight, we see the development of the modern, cash-efficient startup. VCs can make small bets on early-stage companies, and reward those who learn and adapt the most to take into account the information they gain. Customers can get earlier information about what a company’s planning to do, and decide whether that is valuable for them—along with earlier information about when the company decides to pivot away from doing that.
Of course, with hindsight, we also see the seeds of surveillance capitalism and enshittification. When your experiments show that it is easier to grow a free product with ads than a paid-for product. When A/B tests show that people interact more with revenue-generating ads if you design your product in this way.
We also see the era of small bets. If you do not get a result in a couple of weeks, pivot away. Do what customers react well to today, not what they might be absolutely clamouring for in a year’s time, or a decade’s.
It is not the only way to run a business, it is just the only way that gets funded.
Cover photo by Adrian Kosmaczewski.