While pursuing creative endeavours, it’s impossible to follow any fixed plan—be it while writing an essay, or building a startup. Creatives endeavours are bound to be chaotic. There’s a lot of figuring out to be done on the way.

Indie game developer Gabe Cuzzillo wanted to make a stealth game where you had to move silently through corridors and sneak up behind guards—to grab and throw them on walls. By the time the game was completed, it was a completely different one. There was no stealth; not even a human player. Instead it became a smash ‘em up game with an ape as the central character. The grab and throw mechanics became the central theme. It was called Ape Out.

This is a common incident in all creative endeavours. You start with a script or a plan, and soon you see yourself deprioritising them in pursuit of creativity. In game design, this methodology is called following the fun.

Little experiments, random bugs, or unplanned side-effects often become interesting enough to define the central theme. In Ape Out, the grab and throw mechanics added most of the juice to the game for it to take centerstage eventually. Henceforth every design decision was anchored to it. This is what I call the Anchor Mechanics.

Once you figure out what your Anchor Mechanics is, you can start to prioritise ruthlessly. What doesn’t fit can be rejected—be it a game or a business.

Burbn was a mobile check-in app released in 2010. It let users check in at particular locations, make plans for future check-ins, and earn points for hanging out with friends. It wasn’t terribly successfully, and was very similar to Foursquare—another popular mobile check-in app. However, the founders kept tweaking it. They later added a photo-sharing feature to differentiate it from Foursquare.

When they paid attention to how people were using it, they realised that people weren’t using the check-in features at all. What they were using, though, were the app’s photo-sharing features. They were posting and sharing photos like crazy.

At that point, the founders focused solely on the photo-sharing infrastructure—their Anchor Mechanics—and scrapped everything that didn’t fit. Burbn became a simple-photo-sharing app called Instagram.

You may start off with an initial plan, but it isn’t written on stone. Treat it as a guess, not a roadmap. We get 90% of the ideas after we sit down to write. Similarly, we figure out the business model after we start the company.

The game streaming startup Twitch spun out of the general-interest streaming platform Justin.tv. Once the founders realised that majority of users were live streaming video games, it became their Anchor Mechanics. Similarly, Twitter started as a personal status updating platform before focussing on breaking news and entertainment.

But instead of letting things take their natural course, we have a tendency to force our preconceived notions. This is topdown design. Topdown approach starts with a big picture, and breaks down from there into smaller segments. It may bring order and structure, but it hinders creativity.

A bit of chaos is necessary to enhance creativity. This is true for designing a game, writing a story, or running a startup. Innovation has no fixed path—it’s chaotic by definition. Therefore, you have to allow just enough chaos in the process to foster creativity—as long as it doesn’t turn the whole project into a chaotic mess.

Do little experiments. If something works, do more of it. If you follow the fun, or repeat what works, it would feel like the creation is creating itself. It’s an indication you are on the right track. Eventually you’ll stumble upon something that is good enough to become the central theme, or the Anchor Mechanics.

This process is counterintuitive. It’s consciously letting go of things, and deliberately losing grip of your project. But that’s the point. A creation is more about the creation than the creator. Ego makes us feel that it’s about us, but it’s not—be it an art, a story, or a startup. All good creations create themselves. We all are mere experimenters who facilitate the process.

In the 1990s, when two Stanford friends started with a thought experiment of downloading the whole internet, they didn’t set out to build the next Google. They stumbled upon it while trying to solve interesting problems. Getting started with a mindset of figuring things out on the way gives you an unfair advantage—because as others’ plans are failing, you are just getting started.