How Changing Gears Stopped My Startup From Failing For our series The Grind, the founder of Practice Makes Perfect discusses how pivoting helped him not fall into the category of failures.

My startup Practice Makes Perfect -- a nonprofit focused on partnering with schools and operating their summer programs in inner-city neighborhoods -- just celebrated its third birthday, and I couldn't be happier (and relieved) to hit this milestone.

In honor of our third year, I wanted reflect on the instrumental pivots and changes that helped us escape some of the pitfalls that 44 percent of startups face that lead to failure in their first three years.

When my friends and I originally started Practice Makes Perfect at Cornell University, we wanted to give back. Three of the five of us were raised in New York City in low-income households and attended struggling public high schools. It would not be far-fetched to say that the original program design was derived purely from an emotional standpoint. That was okay, as I believe the most effective startups begin because of a problem that the founder feels needs to be solved. Yet, as almost every entrepreneur knows your initial idea is never the end result. To ensure our startup didn't fail, we needed to make pivots to our plan based on feedback, knowledge and variables out of our control.

Related: A Pivot Could Save Your Mobile App From Failing

Figuring out when and why to pivot is difficult. Here are a few insights I learned about changing gears along my entrepreneurial journey.

The gut-based pivot. These pivots are based on the founder or founding team's instincts. They are common earlier in the development of the business model.

When we were first putting together our concept for Practice Makes Perfect, we wanted to pay the high-school mentors and the college interns minimum wage. The cost to operate the model for 30 students would have been almost $50,000. Our inability to raise the funds necessary to execute that model forced us to pivot or live with the idea of not existing. As a result, we found ways to create value in nonmonetary ways by including test prep and college readiness support for our mentors, which allowed us to pay them a more modest stipend.

Research and need-based pivots. These shifts are based on information that already exists.

When we ran our first program in 2011, we were serving 4th grade and 9th grade students. Slowly, the problem of the summer learning loss was gaining greater traction. If we wanted to serve the bigger needs, we realized we would have to find a way to expand our programming or become obsolete. We took our model from 4th and 9th grade to serving students in grades kindergarten through 12th grade.

Experience and evidence-based pivots. These pivots come from piloting and evaluating results.

For two summers, we operated our program using a mentor to mentee ratio of 1:2 without any real evidence to support the group size. Last summer we decided to change things up, so we paired groups from every combination from one mentor for two students to up to six students. We found students' performance was best when the ratio was one mentor for every four mentees.

Related: Do Pivots Matter? Yes, in Almost Every Case.

Feedback-based pivots. These pivots are based on what your stakeholders suggest. Depending on your line of business, time and scope of work you may or may not have the capacity to conduct extensive market surveys to identify your customer and how much they are willing and able to pay for your product or service.

Following our final pilot, we transitioned our organization's business model from one that relied heavily on philanthropy to one that relies on fee-for-service. As such, in trying to lock in the appropriate price point, we had several calls with charter school principals and public-school leaders. Each potential customer provided feedback, and we ultimately decided the direction.

Related: Don't Be a 'Pividiot.' Follow These Tips.