Tackling a hard project? Do this first.

You’ve just been put in charge of a particularly audacious project at work. Your boss says you have to get a monkey to stand on a pedestal and then train it to recite passages from Shakespeare.

How do you begin?

The question comes from Astro Teller, who’s the Captain of Moonshots at Google X (yes, that’s his real title). Now known simply as X, the notoriously secretive company is dedicated to finding radical solutions to huge problems by developing breakthrough technologies like self-driving cars, autonomous drones, and contact lenses that measure glucose levels. X doesn’t innovate for Google. X creates the next Google.

Let’s go back to the monkey-on-a-pedestal exercise. If you’re like most people, you begin with building a pedestal. At some point, “the boss is going to pop by and ask for a status update,” as Teller explains, “and you want to be able to show off something other than a long list of reasons why teaching a monkey to talk is really, really hard.” You’d rather have the boss give you a pat on the back and say, “Hey, nice pedestal, great job!” So you build the pedestal and wait for a Shakespeare-reciting monkey to magically materialize.

But here’s the problem: Building the pedestal is the easiest part. “You can always build the pedestal,” Teller says, but “the risk and the learning comes from the extremely hard work of first training the monkey.” If the project has an Achilles heel—if the monkey can’t be trained to talk, let alone recite Shakespeare—you want to know that up front.

What’s more, the more time you spend building the pedestal, the harder it becomes to walk away from projects that shouldn’t be pursued. This is called the sunk-cost fallacy. Humans find it hard to abandon things if they’ve invested time and money on them. If you spent a bunch of time carving a gorgeous pedestal, you’ll be reluctant to call it quits.

The monkey-first attitude shut down a project called Foghorn at X. The project was promising at first: A member of X read a scientific paper about taking carbon dioxide out of seawater and turning it into affordable, carbon-neutral fuel with the potential to replace gasoline. This technology sounded like something out of a sci-fi movie, so X—true to its form—took it on.

But it turned out that the technology was the pedestal—it was relatively easy to turn seawater into fuel. The monkey was the cost. The process was expensive, particularly in the face of declining gasoline prices. So the team decided to trigger the kill switch and shut down its own project.

I get it: There’s far more certainty in building a pedestal than in getting a monkey to talk. We know how to build pedestals, so we build them. In our lives, we spend our time doing what we know best—writing emails, attending endless meetings—instead of tackling the hardest part of a project.

And it’s not like building pedestals is completely unjustified. After all, the project requires the monkey to stand on a pedestal. Carving the pedestal gives us the satisfaction of “doing something” about the problem and getting some sense of progress—while postponing the inevitable. All this churn feels productive, but it’s not. We’ve built a beautiful pedestal, but the monkey still isn’t talking.

Here’s the thing: What’s easy often isn’t important, and what’s important often isn’t easy.

The next time you’re tempted to build a pedestal, train the monkey first instead.