It is easy to mistake getting things done fast with getting things done well. Growth hacks are more popular than slow builds and unicorns are being built nearly overnight. However, like a bank account, certain things will only expand over time. You can call it the compound interest of entrepreneurship.

Relationships: It takes time to build your social network--not your Twitter followers, but your real connections. Mutual trust, values and goals aren't always immediately apparent. Worse case scenario, you can be so eager to build relationships that you can come off pushing and demanding, creating exactly the opposite effect of what you want.

Slow network building is a skill you will always need as your business evolves. In my case, the social network I built as a journalist needed to be rebuilt when I became an author and, again, when I became an entrepreneur. All this takes time.

Customers: Identifying the wrong ideal customer could be disastrous to your business, as it could have you building an unnecessary product or even realizing (too late) that no one wants what you are offering.

Some of the biggest companies today started with understanding customer needs: Slack originally got funding as a video game company, but the leadership realized there were different customer interests. Compare the $2 billion Slack scenario to Clinkle or Color, two well-funded startups that struggled partially because of an unclear userbase.

Strategy: It's easy to mistake the Lean Startup model for making a quick plan, but it's not the same. If anything, it advocates more time strategizing and less time building. After all, how do you know if your business is hitting its targets, nevertheless if you are creating the business you should be building, if you don't have a well thought out strategy?

And planning doesn't require action, but it does require time. The most thoughtful entrepreneurs take time to think, whether it be walking and shutting down or getting away and contemplating bigger ideas.

What are you rushing that you shouldn't be?