"You need blind faith/no false hope." These two lines from the  Foo Fighters song "Congregation" describe how to launch a successful company. That's according to Andy Jassy, CEO of Amazon since Jeff Bezos stepped down in July, who also says Foo Fighters is one of his favorite bands.

In an onstage interview at the GeekWire Summit this week, Jassy told GeekWire co-founder Todd Bishop why he thinks those words are so powerful. "I think when the Foo Fighters wrote it, they were referring to religion. But I always thought that was such an interesting lyric with respect to new businesses," he said. "When you have a new idea that hasn't been done before, you need that conviction, that blind faith that it could work. And there are going to be a lot of people that will tell you, 'Nah, that's not going to work, people aren't going to care about that. There's a reason it hasn't been done before.' All the naysayers will tell you all the reasons why it's not going to work. You need that blind faith and conviction."

Jassy understands the need for blind faith from personal experience. He led Amazon Web Services (AWS) almost from its inception, which meant trying to persuade CIOs and other leaders from large and small companies to trust their data or operations to Amazon's new cloud service. The team wondered how to convince these leaders that they would have both security and reliability running on AWS. Ultimately, he said, "No matter what we promise customers or prospective customers, you have to prove that over a long period of time."

Which brings us to the second half of Jassy's favorite song lyric--"no false hope." Even though you need blind faith, he said, "you can't just go careening toward your idea without checking in with customers and people who might use it and making sure you're on to something. You need blind faith but no false hope, and every single one of the products I've ever been part of or seen the company work on, you've needed that combination. So I really like that lyric because it's an interesting way to look at business as well."

Ravenous for the truth

In fact, he believes too many company leaders are operating on false hope, even if they don't know it. "I think you need a leadership team that is absolutely ravenous to get at the truth," he said. "I think a lot of leaders have the data obfuscated from them by teams. They think they're doing the leadership a favor sometimes." You have to push past that, Jassy said. "You have to really understand what your customers think about you, not just what you read but get the real data and get a balanced and accurate view."

The reason this is so important, he said, is that for a company to survive for the very long term--such as 100 years--it must be able to reinvent itself. "When you need to make changes, even when people don't want to change," he said, "you have to pick the company up and change it, even if it means cannibalizing yourself." 

That's very hard for most companies to do, he added: "It is really emotionally difficult to cannibalize a business that we've spent a lot of time building and that's doing well."

Jassy experienced this firsthand when Amazon decided to open its platform to third-party sellers. At the time, it was a controversial move, he said. But Amazon's leaders observed the success of online marketplaces such as eBay, and concluded that this was something their customers wanted. "At the end of the day, if customers want something--whether it's good for your business or not--it's going to happen," he said. And so the company created the Amazon Marketplace, which now accounts for about half the retail items sold on Amazon.

The ability to experiment and make big changes even to a business that's doing just fine is something companies must cultivate over time, he added: "You can't wake up at the time that you need to reinvent and decide you're going to do it. You have to build that muscle. You have to have that experimentation model, and you have to have people who are comfortable doing it." Because if you don't, it's highly likely that someone else will.