Amazon is a collection of hundreds of decentralized startups, operating on a set of standardized processes and systems. A core centralized process includes the investment process, or process to allocate headcount to these organizations.
While the exact implementation varies widely across Amazon, a key element is that each organization has a stack rank of how they would spend their investments, starting at zero. In other words, this year your organization may have 81 employees, but that organization will not only need to justify any additional resources, but will need to re-justify those 81 employees as well.
Therefore, each organization needs to prepare a full list of their investments, from basic maintenance to new features. Frugality is a principal deeply embedded in Amazon culture. There's constant pressure to reduce the percentage of resources spent on maintenance and increase the percentage of resources spent on adding value. While some employees mistakenly translate that pressure into a negative statement about quality, the intention and impact is to force the highest customer and business impact from each dollar spent.
Anecdotes regarding resource allocation
The below is a situation I observed a number of times. The Amazon zero based resource allocation process forces a regular look at all projects and businesses.
Darlene submitted her yearly budget for her team. Her group had slowly been entering the mature area of the S-Curve. Growth had been slowing for a few years, and there was no expected acceleration on the horizon. Her organizational plans included maintenance, some operational excellence items, some updates to existing features, and then a few new features.
In her OP1 meeting (yearly planning process), Darlene was told that her organization was going to have its headcount decreased to support only her operational items. All of her new feature proposals were cut. Those new features were not going to change the trajectory of the overall business, so they weren't a good use of Amazon's money. Unless they proposed new projects to reinvent her business, it would stay on life support.
This was naturally distressing to Darlene and her team, who preferred to continue to iterate on their product rather than simply support it.
The below is a mirror image of the above. In this case, the ability to move resources is key to how Amazon can accomplish great things.
Santiago proposed a roadmap for his product to Jeff Bezos and his staff. Jeff said he was excited about the roadmap, and offered some feedback. Towards the end of the meeting, Jeff said he agreed that the team needed to move forward. Except that the 3 year roadmap would need to be changed to launch in 9 months.
Santiago objected, saying that the team couldn't deliver a 3 year roadmap in 9 months. Everyone knows that Jeff doesn't like "can't". He insisted that Santiago's team find a path, or come back with a convincing argument that it was literally impossible.
Santiago talked with his team, and came up with a proposed hiring ramp and roadmap which would achieve the launch. They assumed this was a wasted effort. The updated roadmap included eight times the headcount of the previous budget, and required a movement of existing employees from other groups to the team immediately.
Santiago met with Jeff again, expecting Jeff to object to a number of his proposals. Instead, the document was read, Jeff nodded and asked them to go execute.
Santiago and his group was left stunned. For better or worse, their project was now fully funded, and it was time to go deliver.
Resources before roadmap
Many companies operate by having each organization have their set amount of resources, perhaps increased based on that organization's priority. Then each organization determines how much they can accomplish with the available resources.
This allows an organization or group to work independently spending their money. However, this decreases the chance that an individual organization will make the hard choice of cutting investments if they aren't globally the best use of resources. This also decreases the chance that an individual organization could make a massive investment in a particular area of opportunity.
If existing groups own their resources and allocate them as needed, you would expect as each group matured, that they would spend a higher percentage of their funds on maintenance rather than new initiatives.
In aggregate, this means as a company matures, their overall percentage of investments in innovative projects decreases, as mature businesses own the majority of resources. While these businesses are not necessarily shrinking, they are not growing at the rate of new initiatives, new business lines, or new opportunities. They are not what defines a successful company.