Amazon's Panos Panay: A $1 Billion Write-Down Taught Me Everything

A nearly $1 billion write-down on the Surface RT was, by Panos Panay's own account, the best education he ever got. Amazon's SVP of Devices, Alexa, and Leo laid out that philosophy on stage at TechCrunch Disrupt in San Francisco, telling startup founders that failure isn't optional — it's required.

Panay spent more than 19 years at Microsoft before joining Amazon. He weathered one particularly expensive flop there: the Surface RT, Microsoft's first tablet computer, which launched incomplete and ended in a massive write-down.

"You have to be willing to fail, and actually, not just willing, you actually have to go fail," Panay said. "If you haven't failed, you probably haven't pushed the boundary. It's literally about harvesting the lessons learned when you fail. You have to be willing to face it, and as you face it, learn from it."

What did the Surface RT failure actually teach him?

Panay didn't soften the memory. "I remember being on the front cover of a newspaper, and it just said failure," he said.

But he argues the admission was the easy part. The hard part — and the part that matters — is the forensic work that follows.

"The trick was, and always is, did you harvest the learnings and understand what happened?" Panay said. "Because the failure wasn't that moment. It was decisions that led up to it two years before that, whether the software wasn't complete or the inventory was wrong or how we balanced into retail. There were all these steps along the way."

His diagnosis of the Surface RT collapse, in his own framing:

• Incomplete software at launch
• Inventory planning that missed demand
• A retail strategy that didn't balance the product correctly

Each decision sat two years upstream of the write-down. That's the timeline founders should audit, he argues — not the launch-day headline.

The takeaway in plain words: the failure you see is never the failure that happened.

How should founders respond when it happens?

Panay's prescription has three parts. Accept the failure without excuses. Dig into what went wrong. Then hold the vision steady even as the execution changes.

"Here's the other side of it," he said. "You want to succeed, yeah? You take that failure, but you don't let go of your vision. That's it. Surface has done pretty well. You don't let go of your vision. You don't break it."

The Surface line's eventual recovery is his proof of concept. The product that generated a roughly $1 billion write-down became a franchise Microsoft kept building for over a decade.

The takeaway: absorb the hit, keep the destination, revise the route.

How much should a founder pivot after failure?

Not much, in Panay's arithmetic. He puts the right adjustment at roughly 30 degrees, not 180.

"It goes back to the same fundamental, like, have you learned?" he said. "Were you learning along the way? Because if you were learning, then you just change your vector. Usually, it's not 180 degrees. Usually, it's just 30 degrees. Take your lessons learned, change a little, adapt your vision where it needs to be, and then go."

He drew a sharp line between persistence and denial. "I believe there's people that hang on too long, but it's because they haven't accepted the moment of failure at all," Panay said. Founders who refuse the initial admission end up trapped; founders who accept it can adjust fast and move.

The takeaway: a small, informed course correction beats a dramatic one.

Why this matters for Amazon's own bets

Panay now runs Amazon's devices portfolio, including Alexa and Leo — product lines that live or die on iteration. His message to founders doubles as a description of how he operates: move quickly, learn from setbacks, adapt the vision where needed, and keep shipping.

The so-what for operators: the next product failure may cost less than the lessons it hands you — provided you actually go harvest them.