Subscribe to Entrepreneur for $5

Iterate or Eliminate? When You Need to Go Back to the Drawing Board.

Opinions expressed by Entrepreneur contributors are their own.

There's this myth that entrepreneurs have a eureka moment on which they build a business and get rich. This idea couldn't be further from the truth.

Startup Stock Photos

The unsexy truth of building a business is that the idea that inspires an entrepreneurial path often becomes irrelevant in the reality of the marketplace. Entrepreneurs are continually pivoting to ensure a strong product-market fit.

The ability to bob and weave is ultimately what counts. This is the central idea behind lean startup principles, which "favors experimentation over elaborate planning, customer feedback over intuition and iterative design over traditional "big design up front' development," Steve Blank, one of the key evangelists of the "lean startup" movement, wrote in the Harvard Business Review.

For bootstrapped startups like NerdWallet, this puts a name to what has long been a reality. Flexibility is encoded into a bootstrapped company's DNA -- you can't waste too much time on an unproven idea that isn't generating cash.

Related: 6 Signs It's Time to Turn Your Startup in a New Direction

As 75 percent of all venture-backed startups fail, according to Harvard research, time is short to prove your idea or move on to something better.

Here are the keys to know when to double down, when to iterate and when to let go.

Your first idea probably won't be the end result. When a group of Vancouver developers worked on an online game, as an afterthought they added a photo-sharing tool for players. You've probably never heard of the title they were working on, "Game Neverending," but I'll bet you've heard of its photo-sharing function: Flickr.

This is a common tale among successful ventures. Twitter started as a side project at a podcasting company as a way to send SMS messages over the Internet to a small group. New media company Vocativ was built on the back of deep-web search tools developed as security software for governments and corporations. Don't be so blinkered by your baby that you miss where real opportunities lie.

Don't lie to yourself. Entrepreneurs are very passionate people. Once you come up with an idea that feels right, you can become consumed by it: It takes over your conversations by day and your dreams at night. That zeal can be key when weathering the inevitable storms you'll face when bringing your product to life. But it can also be a liability if you ignore evidence that your great idea isn't so great.

You have to pause and ask yourself: Am I being honest with myself? Am I dwelling on self-verifying facts that confirm my original hypothesis, no matter how slight? You need the dispassionate approach of a scientist testing a hypothesis to give frequent reality checks to your enthusiasm.

Related: A Pivot Could Save Your Mobile App From Failing

Being honest with yourself is hard, but it's much harder if you have investors: You've sold them on an idea, which makes it that much harder to go back to them, hat in hand and tell them you've changed your mind.

Experiment and iterate. For technology ventures, advances in Web programming languages and a whole slew of services make real-time testing of an idea easier than ever.

At NerdWallet, we make a practice of minimum viable product (MVP) development. We're not alone. When Dropbox was trying to get user feedback on its development, founder Drew Houston did something "unexpectedly easy," as Eric Ries writes. He shot a video.

"The video is banal, a simple three-minute demonstration of the technology as it is meant to work, but it was targeted at a community of technology early adopters," Ries said. "But the beta testing waiting list went from 5,000 to 75,000 literally overnight."

Online shoe store Zappos began with a "Wizard of Oz" approach to MVP, making customers think the process was automated when in fact behind the curtain employees were running to shoe stores. The data collected through that experimentation helped build automation down the line.

Similarly, Food on the Table built a scalable business by manually buying groceries based on a set menu for customers before automating the process. Their lesson: "Learn first, code last."

Make improvements quickly. There are always going to be borderline decisions. Things are moving in the right direction but are they heading there fast enough? Maybe you make a tweak that turns up the conversion rate of sales 10 percent -- that's a solid improvement. But maybe you need something like a 50 percent improvement in conversion rate to make it worth its while.

Related: Trim the Fat From Your Startup

How much more time you spend deciding whether to tinker or cut loose is more art than science. But beware the lure of modestly promising results when you're looking for a rainmaker.

Let it go and pay your "tuition." If you find yourself doubling down on an idea in part because of all the time and resources you've devoted to the project, that's a big red flag that it's probably time to let it go.

"But I've spent months spinning tons of line of code," you may tell yourself. "If I just change this one thing, I know it will work. I don't want all this to go to waste."

But it wasn't for nothing -- you learned what didn't work. I used to feel like I'd wasted a lot of time or money when results disappointed. Now I view those experiences as expensive tuition in my entrepreneurship degree. Pay your dues and move on.

Related: How Changing Gears Stopped My Startup From Failing

Entrepreneur Editors' Picks