Short answer – almost all the Unicorns pivoted. The authors of the article didn’t understand what a pivot was.
What was lacking in the article was a clear definition of a pivot. A pivot is not just changing the product. A pivot can change any of nine different things in your business model. A pivot may mean you changed your customer segment, your channel, revenue model/pricing, resources, activities, costs, partners, customer acquisition – lots of other things than just the product.
The 9 different things, in this case, are the 9 boxes in the business model canvas.
However, I’d push this a bit further. Once upon a time, source code control with solid tools like Subversion meant branching was difficult and required agreement between multiple people, etc. Then Git and Mercurial came along, and now branching is free, something that happens constantly, smoothly, without overhead. Best development practice has evolved accordingly to encourage frequent branching and merging.
Once upon a time, making changes to the business model was poorly understood, something you did by gut feeling, which had a high risk overhead because there were no measurements to prove that a change was beneficial. With Lean Startup and the Business Model Canvas, or other, proper mapping of your assumptions (such as Hypothesis Driven Development, making changes is something that can happen smoothly, quickly, and on a daily basis (at least pre-market-fit). Until you get to product-market fit, your main advantage is your speed of adaptation.
Like continuous deployment, which involves deploying the codebase to production many times a day, each time a new change is made (and tested), continuous pivoting means making tweaks to (often small aspects of) the business model several times a day. Most of those “pivots” are in fact micro-pivots, but in aggregate they can add up to very significant changes. This makes obvious sense from a pre-fit perspective, but what about post-fit?
I would suggest there is no fundamental difference between a pre-fit business model tweak and a post-fit one. And as the environment that the company evolves, the company must adapt – through continually pivoting. It might be called something else, but that’s what it is.
These changes necessarily become slower as the company grows and develops more cultural inertia, and as the risk of uncontrolled changes grows, but when they stop, or become too slow to adapt to a changing market, the company stops growing and starts dying.
Which leads to another interesting insight: once your startup grows, if you want it to live on for a longer time, one of your primary concerns as a founder must be to build a human organisation that has the nimbleness necessary to keep pivoting, quickly, forever.