How Startups Turn a Three-Year Goal Into a 90-Day Plan

Too much long-term vision can lead to short-term failure

Photo: Jon Feingersh Photography Inc./Getty Images

Let me ask you a strategic question: Where do you want your company to be three years from now?

• What will your product or service look like? Which new features will you roll out, and how will your business evolve?
• How will your market share expand? Which new markets will you have entered?
• How many customers will you have? How much revenue will they generate? How many employees will you need to support them?

Got all that? Great. You’ve got 90 days.

The future is NOT now

One of the hardest things I have to do, either as a leader or an adviser, is to get my best people to stop thinking three years out and start thinking about today. Don’t get me wrong — these are my best people because they have vision, because they can see the future and are constantly working toward it.

But sometimes these same folks will take a good thing too far. They Luke Skywalker that shit — they can’t stop looking to the future or the horizon, their minds never on where they are and what they’re doing.

I know this type. I am this type. I’ve been in the startup world for over 20 years, and if I look at my failures and my low points, in startup or any other pursuit, the reason I failed wasn’t that I was doing too little; it was that I was doing too much.

The dark side of vision is future-creep

Entrepreneurial vision is a necessary and often overlooked skill. It isn’t about riding trends or wielding psychic magic, it’s about being able to look at patterns today and visualizing tomorrow’s results. In more concrete terms, it’s the opposite of that one dude who just comes in and does his job every day.

But too much vision always leads to future-creep.

With early founders, future-creep usually manifests itself in the overbuilding of infrastructure. I have artifacts to remind me that I’m prone to this behavior, like boxes of old business cards from my previous startups (both successful and unsuccessful). I also have old code, old…