This post was originally published on LinkedIn.

I recently had the privilege of attending the Sun Valley conference, where I listened to some very creative people discuss their innovative ideas for progress into the future.

One of the most interesting was Elon Musk, founder of Space X, talking about the challenges of running a company that designs, manufactures, and launches advanced rockets and spacecraft.

His company had a huge blowup last month when a rocket exploded shortly after lifting off for the International Space Station. He talked about all the intricacies of running such a complex operation, all the work that’s involved in making sure there isn’t a problem. And then, what happens when – even after all that careful planning – a rocket still blows up.

The challenges he faces reminded me a lot of schooling. A school, like a rocket, is a delicate mechanism. So many things can go wrong when teaching children, just as when building and launching a rocket. But one advantage his company has is that his rockets don’t have any people on them. Schools do.

When a class blows up, or when a school fails, real kids are hurt. Not physically, but their minds are hurt. Their futures are hurt.

That is why I strongly believe in making changes right away if something is going wrong. Schools often take the view that if they have a problem, fixing it should wait until the next year. I don’t believe that. If you waste a year of a kid’s life, the child will never get that time back. A student can’t just start over the way you can build a new rocket if the old one explodes.

In building Success Academy, we had to go back and retool more than once, as our carefully planned design didn’t work out as expected. Some of these changes were major and foundational; others were small but had big implications for how the schools operated.

For example, we started Success Academy using a very popular and well-regarded math curriculum. But we discovered that instead of moving the children quickly toward a deep understanding of mathematical concepts, the curriculum was wasting their time. Kids were spending three weeks on the number 7. It was horrifying! We had to rapidly rethink our approach to math and put in place a completely different program that was far more conceptual and rigorous. It was disorienting for the teachers and the scholars, but it was the right move.

We had the same problem with the literacy curriculum we started with – everyone thought it was great, but we found it wasn’t nearly good enough. In fact, none of the popular reading and writing curriculums set the bar high enough, so we wrote our own: THINK Literacy, which draws on the balanced literacy approach but places an extraordinary emphasis on content over procedure. Even in kindergarten, children think critically about what they’re reading and debate the meaning of the text.

We also didn’t understand at the beginning that if principals were to focus primarily on instruction, they needed to be freed from the burden of running their buildings. It made sense to have principals be in charge of everything that happened within their schools, but if they had to deal with all the government compliance that charter schools have to go through, they’d never have time to get into a classroom! That’s why we put a business operations manager in every school – to free up principals to deal with the most important work: teaching and learning.

Besides curricular and managerial missteps, we had some procedural problems as well.

We felt, and we still feel, that it is important for the children to wear uniforms. For the boys, this means wearing a tie. But we don’t give ties to our kindergartners anymore, because the boys kept sucking on the ties and swinging them around. The teachers were spending way too much time taking the ties out of the boys’ mouths. So we ditched them.

By learning from our mistakes, constantly reassessing, and fixing problems now, not later, we built a culture of success that can persist even when something major – like a rocket – blows up.