We've talked to some people at Amazon about the culture, and I think it's an important aspect. How would you characterize the culture? Specifically, I'm curious about how you work with the finance team to understand how these innovations and efficiencies affect the P&L. It's a two-part question around, firstly, the culture of like how people at the company think and how they're motivated and what they're hoping to do. And then, ultimately, where the rubber meets the road. How are you working with the finance team where maybe, on one hand, you have like these big audacious ideas, on the other hand, like it has to result in an improvement.

We prioritized the largest cost segments. For instance, the acquisition of Kiva was based on the fact that roughly 50% to 60% of the labor costs in a warehouse were due to the time spent walking to pick up inventory. Automating this was the first step. Other steps, like item manipulation, involve multiple touches within the warehouse. Many technologies Amazon has introduced aim to reduce these touches, thus lowering variable costs.

We've talked to some people at Amazon about the culture, and I think it's an important aspect. How would you characterize the culture? Specifically, I'm curious about how you work with the finance team to understand how these innovations and efficiencies affect the P&L. It's a two-part question around, firstly, the culture of like how people at the company think and how they're motivated and what they're hoping to do. And then, ultimately, where the rubber meets the road. How are you working with the finance team where maybe, on one hand, you have like these big audacious ideas, on the other hand, like it has to result in an improvement.

Proteus is another example, doing for indirect labor costs what Kiva bots did for inventory—moving carts or resupplying workstations. The approach was methodical of, from a business perspective, you're running the P&L and looking at the cost to fulfill an item. What are the different pieces that step into that total amount that it's costing the company? And let's start chipping away at the biggest ones first, using technologies that we either have today, create in the near term or that we can invest in for the future.

Is that also a reflection of some of the efficiencies within the manufacturing of the robots themselves, where the prices for integrating robotics have declined, making the economics of labor versus automation more attractive?

Yes, but not really in the case of Amazon. The reason is that if you look at some of the larger fulfillment centers, they can easily fulfill 250 million packages or items a year. When automation touches 250 million items annually, you quickly amortize the capital costs. It's not the cost of the robotics themselves that becomes the biggest impediment to making those investments. It's more about the time and what you're unlocking within the warehouse. So, for Amazon, that's not really the case. However, in other parts of the industry outside of Amazon, that is certainly the case.