The Amazon Management System

Last Updated on 3 months by Piyush Jain


The Amazon Management System

By Ram Charan and Julia Yang

Book Review

The book is one of the knowledge gems from Ram Charan. Reading the book, one feels that these things are quite obvious while managing an organization, but the alternative approach and implications are also shared for good results.

The book clearly states that the times have changed, and primitive methods of management and growth won’t apply anymore. We are talking of self-growing, ever-evolving, innovative, self-managed, and result-driven organizations.

Amazon is one of the best examples where a company evolved from basic business models to utilize the latest technology and innovative practice while remaining big, and still be competitive enough. Most of us being Amazon application users (shopping, fire stick, audible, music), this story goes well in our conscience without much struggle.

In the summary of the book, I aim to give some points with respect to Amazon Management System which are above the subjective statements, so that the audience can co-relate with their working environment. As Amazon focuses mainly on the B2C segment, it mainly focuses on the customers, while the other companies which are in other areas, can also correlate with their reason of existence, vision, and mission.

The book does glorify Jeff, but that does match with the growth of the company and the innovation the company has done for society in a positive manner. Jeff, no doubt is a tough taskmaster with strong values and principles.

The authors have divided the Amazon Management System philosophy into the 6 building blocks.

Amazon Management System Building Block 1:

Customer Obsessed Business Model

This business model is more about seeing the business model as not generating money as a primary objective, but serving the customer and getting money as a byproduct. When Jeff was with DE Shaw on the Wall-street in 1994, he brain-stormed with the CEO about the possibilities the internet can increase for the customers. They analyzed that the basic number was 2300% growth in the business, which leverages the internet for customer reach.

What drives the business. Jeff figured out that the customer is the ultimate driver, and his demand can drive your growth. As Jeff states, “One thing I love about customers is that they are divinely discontent. Their expectations are never static – they go up. It’s human nature.” No competition, product knowledge, or only technology can drive the business. To gain customer’s trust, one focuses on the customer requirements, pains, and problems, and works around them. The only way to do this is through continuous innovation and relentless invention. In this way, the divinely discontent customers became the sources of continuous inspiration for Amazon’s invention machine. One has to think innovatively to solve the problems. At Amazon, any instance of customer dissatisfaction is dealt with as an internal severity and is resolved for maintaining customer trust.

Long term thinking – Along with customer demand, Amazon promotes long term thinking. They developed AWS, Echo, Alexa, Fire Mobile, and many others. Amazon went ahead with speed and scale to implement the idea. This way they would learn fast and fail fast, without burning more time and effort.

Amazon Management System Building Block 2:

Continuous Bar-Raising Talent Pool

To bring all these plans to reality, Amazon needed the best talent that would fit their culture. Amazon is looking for builders. People who can think, do things, take responsibilities. Any wrong hire can lead to more pains, revenue loss, and emotional distress.

The authors give a very good example of Zappos, where the company offers monetary benefits to the people who want to quit while saying that they do not want the people to accept the offer. This way, while they have given the offer, they bait people who do not want to be there to accept the bait. Amazon also followed the same policy and weed out the disengaged employees.

Furthermore, Amazon wants to have people who are leaders and they have a sense of ownership. They act on behalf of the entire company, beyond just their team. They never say, “that’s not my job.”

Amazon does not want its employees to behave like tenants. Few characteristics of a tenant would be like

  1. Job insecurity – If I hire a person better than me, then I may lose my job.
  2. Frugality – Why should I look to save money? If I don’t spend, someone else will.
  3. Deep-dive – Why go into detail? This is not my job. If something goes wrong, it is not my problem.
  4. Have guts to disagree and commit. Many people do not have the guts to disagree with their boss, as it may be detrimental to their relations and career progression.
  5. Deliver results. People should be ready to commit and deliver results. Employees sometimes see this as not their job, but their bosses’ as he gets the higher pay and benefits, so why stick his own neck in it.

The person should be mentally strong to be ready to take on challenges, persist hard work on them relentlessly, be ready for failures, and even face rants. Jeff himself walked the talk. He demonstrated extreme mental toughness when he put a deadline on himself to launch “Amazon Prime” in flat 111 days. He conducted 60 meetings for the viability of Kindle and raise the 1st $1 million for the program.

Amazon has a concept of bar-raisers. The bar-raisers aim to hire talent which is better than the previous one. They are away from business pressure and do not compromise on the right fit for Amazon. The bar-raiser has the ultimate veto-power after all the rounds, and scrutiny.

What do the panel members check for? 

1) Leadership principles the candidate followed during his/her career.

2) Failures – The candidate must have done something which led to his/her failure. What did they learn from it, and how they think they can improve on it? Failure is important for people to learn, and innovate.

3) Writing – As Amazon has banned Powerpoints, they aim that the candidate should know how to write, and candidly explain his thoughts in the right amount of words.

The interviewers are trained, given continuous feedback on how to improve interviewing skills and get the right fit for Amazon.

In terms of salary, Amazon is one of the most frugal companies of the big ones. Even to the extent that the employees have to pay for their parking. Bonuses are banned at Amazon. The employees get the shares allocated in the pattern of 5, 15, and 20 percent in the 1st, 2nd, 3rd, and subsequent years. This will ensure that only loyal employees get the benefit of this allocation.

Amazon aims to keep its employees with them for a long. They did not let bureaucracy get in and steal the independence of the talented people. Amazon’s top hierarchy team has remained with them for more than 20 years. When the right person is spotted, they go to any extent to get the person on board.

Amazon Management System Building Block 3:

AP Powered Data And Metrics System

At Amazon, when Bezos throws a question at you, there is no wriggle room for obfuscation or buzzwords. If you dare to try, he will rant at you, ruthlessly, “the answer starts with a number!”

It is all numbers at Amazon. Amazon does not believe at micromanaging anyone, but still keeps full control on the progress of each practice. Number of metrics kept for any practice is humongously high. While in a normal course of any kind of operations we would track 15 to 20 metrics, Amazon has an order of 200+ metrics. Just the list of metrics reached 70 pages in some cases.

But why do other organizations hide information? Information is used as the basis of power, or hidden for various reasons. Confidentiality, lack of authorization, reluctance, personal grudge, delays, distortion, or purposeful omission of information may be few of them. The first rule for digitalization is no hoarding of data. In the customer obsessed business model, all the data belongs to the whole company and no individual.

The aim of metrics tracking is to enable end-to-end accountability and promote transparency.

The executive review may be there for Q, H, or Y, but it is expected that if there is a dip in some metrics, and the review is later, then it is expected that it be looked at within an actionable time-period. Some metrics may even be so important that if their dip is not attended to immediately, it may lead to a big loss to the company or even closure of the initiative/company.

For example,

  1. Dip in page load time by 0.1 seconds can reduce customer activity by 1 percent. Further, you can co-relate.
  2. A decrease in the number of customers contact by 90% can significantly increase profits as less of the support staff is required.

Interesting anecdote – In the year 2000, at the meeting of all the heads of departments with Jeff, Jeff asked the customer service head what is the wait time for the customer on the call. He promptly answered that it is under a minute. Jeff immediately called customer service and started looking at his watch. A minute passed, 2 minutes passed, ultimately the support executive picked up the call in 270 seconds (4.5 minutes). Way beyond the initial claim. Amazon maintains a philosophy, “In God we trust, all others must bring data.”, as said by Edwards Deming. 

Does collecting a lot of metrics become a problem for anyone? No. Metrics are collected, but they do not drop the Day 1 philosophy for their builders. AI/ML systems in the ecosystems, keep churning the metrics, for easy interpretation. This helps reduce hierarchy. While the hierarchical traditional model only allows a person to supervise 6-10 people at a time, but when metrics speak and all can see them, this number increases. The hierarchy in an organization as big as Amazon may increase to 10 or more, while Amazon aims to decrease it to 3 with the use of metrics model with AI/ML. The book gives an example, that Jeff Wilke, a CEO of a Worldwide Consumer, could personally manage 500 project teams at the same time.

The metrics system could also analyze historical trends, and give the outliers for the executives to act, instead of burying their heads in the metrics always, and give time for them to do something innovative. Metrics also help analyze decisions with time. In case, some decision is taken and one does not agree with the decision taken. He/She can easily prove the same using metrics logically and give an alternative course as well.

This automatic monitoring of metrics is the Amazon’s secret ingredient to speed and agility they hold.

How are we sure that concentrating so much on metrics is making a difference?

If we see this space, many other e-commerce giants have still not been able to implement premium programs inspite of so much effort. Amazon has a internal software called “Mechanical Sensei”, This is the answer. It gobles up the metrics, and comes up with the findings. Amazon wants its builders to look at any and all the algorithms in the space, and extract benefit out of it. This can impact anything. Pricing, product categories, sales trends, internal management, or anything under their sun.

Amazon have used few of these algos to understand pricing of the other players. They used this to give the best to their customers. One of the example has been, about the top 100 best selling books. Comparing with other players at some point of time, in 72 were better, 25 were at same level and 3 were below the other players. Why 3 were below other players? It was evaluated, and found that they were out of stock there, so do they let it remain? NO, it has to be improved. Customer should not feel that since it is out of stock on other side, I have to pay a premium to get it on Amazon.

Amazon Management System Building Block 4:

Ground Breaking Invention Machine

All this started from Hoover’s Vaccum Cleaner converted to a primitive hovercraft when Jeff was a child. He wants to inject this DNA in the company he formed.

The company focuses on promoting the use of AI/ML in various spheres of life. It sponsors and participates in various AI/ML initiatives, as well as internally sees that where it can use it.

Any innovation exercise will carry cost and effort, which Amazon boldly proceeds with. Kindle was the first hardware initiative, and before this Amazon never ventured in any hardware initiative of any kind. When Amazon started, none of the initiatives, like cloud, hardware, voice recognition, or AI was in the portfolio of the company.

With Kindle in place, Amazon got insight about the cloud, voice recognition, and release of Echo. Now Amazon’s steps were more grounded, and this step was a “Compounding Effect Of Inventions” for the company. 

Amazon even went to an extent of killing its own business. It says that if you dont kill your own business then someone else will, and then you will not be ready for it. When Kindle was introduced, Jeff knew that selling physical books is their bread-and-butter. If kindle picks up then people will not buy physical books. But if someone else comes up with an alternative, then Amazon will not be ready for it. Best example of this is/was Kodak. Kodak even though it invented digital camera, it did not promote it and filed for bankruptcy in 2012.

Amazon also realizes that as it grows, the failures also will be bigger. Failure and inventions are twins. They go hand-in-hand. If you are going ahead with something new, and you know it will work, then it is not an invention or experiment. Amazon experienced this multiple times in 25 years. They failed big 18 times.

BUT, this failure was important and the learnings were used in the successes of Echo and Alexa. Other companies who are high on innovation, not only seek to invent new things but also see how much they are failing. Netflix CEO Reed Hastings tells that he is not happy as they are not failing enough, which means one is not taking enough risks.

Did success come easy? Not at all. It took lot of effort, time and patience to release successfull products. 2 years for 1st service of AWS, 3 years for Kindle, 4 years for Echo, and multiple years for Amazon Aurora database. But when these services launched they payed of handsomely, and got right benefits for their clients too.

When Amazon seeks ideas, you don’t have to be a top executive or a rich man. Anyone can give ideas. They have an idea cell, which keeps on accepting ideas from all in the company. Jeff also personally looks at the ideas, and if something clicks with him, he will go ahead with it. Junior Software engineer Charlie Ward gave the idea of Amazon Prime, and it was accepted. It gave Amazon a benefit of 100 Million prime members by 2018, along with prime subscription revenue only next to Netflix.

“Simple and Scalable” is the expectation of the idea. His first focus is the customer, and not the competition. If the competition has developed something, it is ok. That does not lure him until it is the demand from the customer. Jeff believes that ideas should be simple to implement and scalable. He also accepts ideas from his customers, vendors, and whomsoever. 

Any innovation effort need to start with customer and end with customer. These ideas should not be like low hanging fruits. They need to be difficult problems to solve but implemented as such that they become easy for customer to achieve the motive.

Who would implement the idea? Obviously Amazon employees, but how many? They call it 2PT (2 pizza team), means the team is only that big that can finish their lunch in 2 pizzas pies. This is a typical size of 6 to 10 people, but motivated.

Who would lead the team? A person who can go to any extent to achieve the goal. He does not give excuses, be dependent on other people, or get stuck in limitations. This person would land in the S Team of Bezos. Truly said, nothing overcomes the wrong person. In the wrong hands, great ideas will not blossom.

Who has to ensure the success of the program? Project team, implementation team, sales team, R & D Team? At Amazon, it is the Project team. At Amazon, the project team is held accountable end-to-end (e2e), meaning their ownership extends all the way from concept, to design, to development, to testing, to launch, and to post-launch operation. Why? Bezos believes in the principle of “Eat your own dog food”. Implement what you have built, so that there is no finger-pointing, or hiding behind someone, or blaming someone else.

Who and how is the decision taken? For that, please read the next chapter.

Amazon Management System Building Block 5:

High Velocity and High Quality Decision Making

Decision is something we live with. It is not the life or the event that decides our life, it is the decision. (This is my belief)

How does it work at Amazon?

Decisions are classified in 2 types.

  1. Type 1 – This is a decision that is a one-way street. That means you have done something, it impacts current systems and there is no point of return if the decision backfires. It is a costly affair as it impacts the running business. One takes this type of decision with the utmost care, a lot of precautions, and analysis at each step.
  2. Type 2 – This type of decision is about the new ventures, and does not impact anyone. This is about experimenting, exploring, and doing new things. Here too much thought and bureaucracy can scuttle the innovative process.

What does it take to take decision?

General Collin Powell states that amidst war if you have information between 40 to 70 %, go ahead, decide on basis of your gut feel. At Amazon, it is 70-90 % range. If you have around 70% information, then go ahead. It is not possible that you will have all the information always. One needs to be observant and be ready for course correction if they feel something is done wrong. Be ready to correct yourself and that will be less costly than taking corrective actions later.

Where does this 40 and 70 percent information come from?

Metrics. Amazon internally has so much historical data that they use that. Or get from a different source. For any decision taken, the result would hover somewhere near the previous results.

What happens when there are cross functional teams?

They just sit together and take decision to move ahead. They will not be entangled in bureaucratic procedures.

So who is ultimately accountable?

For type 1 decision, the senior executives, as they have gone through all the metrics, and understand the market. In some case, even Jeff is ready to take that responsibility.

In general, The project sponsor.

What do Amazon expect from its associates while taking decisions?

  1. Find the best truth – Use metrics to hold your opinion, challenge someone with it, without the fear of hierarchy, command, and control.
  2. Imagine the possible change – While Jeff Bezos wanted to implement Amazon Prime, other executive Brad Stone suggested that it is a bad idea. During the promotion time of 2-day free shipping, Amazon finances go for a toss, so what will happen when Prime is in place? Jeff states that when people buy more, Amazon will have to improve its shipping practice as well as the increased volume will decrease its shipping cost. Furthermore, later the analysis proved that Prime customers only shopped 2.7 times more than normal customers.
  3. Combat Group Thinking – Do not worry about the hierarchy, challenge the point if you think it can be improved or have to be changed. In many places, opinions are not aired to maintain harmony or prevent personal grudge against anyone. But you are doing injustice to the customers and the shareholders in that case.
  4. Have the backbone to disagree and commit – When you have a different opinion with someone, be ready to disagree with him, and prove your point.
  5. Minimize regrets – When nothing works, opinions vary, data is not there, nothing is clear, then go ahead with your gut feeling, so that you do not have regrets later on.

Furthermore, any decision should comply with the 9 guidelines of decision making at Amazon.

How do people conduct meetings at Amazon?

Amazon has banned the concept of powerpoint presentation. They believe that

  1. Points stated in PPT are clear to the presenter, but not to the viewer.
  2. When the presenter starts with the presentation, he states points that are going to come later early, so this creates confusion.
  3. The answers which the viewers may be looking for earlier maybe later in the slides.
  4. PPT does not interconnect the ideas well.
  5. Someone can interrupt you between the presentation and disrupt the whole objective of the meeting.

Jeff states that “When you have to write your ideas out in complete sentences and complete paragraphs, it forces a deeper clarity of thinking.” So, if you read the whole six-page memo, it often happens to me, I get to page two and I have a question. I jot it in the margin and, by the time I get to page four, the question has been answered so I can just cross it off. It saves a lot of time.

When the meeting starts, everyone first spends 15 minutes or so reading the memo and understanding the details. When all are clear, they start the meeting.

So finally when decision is in place, how do one build it? 

For the same, look at the summary of building block.

Amazon Management System Building Block 6:

Forever-Day-1-Culture

What is the Day-1 culture? When a startup is formed, there is no limitation, on who will do what. All aim for the results and do whatever is needed to achieve that. Check out here.

What do they do, so that they does not slip in Day-2 culture?

• True customer obsession – Any effort, starts and ends with the customer in mind. Customer’s divine discontent will help one work patiently, accept failures, and be happier when the customer is happy. This will keep one going relentlessly.

• Resist proxies – Do not offload your work to someone else. This is a common phenomenon when the company becomes big. If you can do it, then go ahead and get it done. Avoid proxies. If there is an unavoidable dependency, then prepare a fallback plan to proceed ahead.

• Embrace external trends – Be ready to embrace new trends. If new technology is there in the market, jump on it before anyone else does, and find how it could be of any use for your advantage.

• High-velocity decision making – Be quick in making decisions. Any slowness will push you to Day-2 culture, where the company will wait to die slowly. For any delay in this, the usual suspects are complacency, bureaucracy, and interdependency that blurs the lines of accountability.

How to kill bearucracy in the organization?

  1. Budget control – No one should aim to build an empire in Amazon. There is a shoestring budget and only required people will be there.
  2. Indirect Headcount – Reduce the number of indirect headcounts. Direct headcounts are technology and customer-facing people skilled. Indirect ones are those in accounting, man-management. Amazon has only 1.5 percent of people who are of indirect headcount.
  3. Simplify the process: Rule should be made to avoid bureaucracy. Jeff makes it clear that the rules should be able to clear the following checklist, made by Jeff Rosemann.
    1. When the rules can’t be explained;
    2. When they don’t favor the customer;
    3. When you can’t get redress from a higher authority
    4. When you can’t get an answer to a reasonable question
    5. When there is no service-level agreement or guaranteed response time built into the process

How does Amazon Management System enforce Day-1 culture in their working?

Amazon Management System enforces the culture in various manners. They have high standards of implementation. Everybody thinks from a customer perspective and delivers results accordingly. Everybody retrospects periodically that can they do something better at this point in time. As a token of respect, they leave 1 chair empty in a meeting room, which is a symbolic gesture for the missing customer/stakeholder in the meeting. To constantly motivate themselves, there are regular press releases of the AI/ML or any new work they are doing, to broadcast to the world.

The executives also have to do 2-day training at the call center to understand the pain of the customer and think accordingly. They constantly look for negative feedback and work upon it to correct it for winning back customer trust. It may happen sometimes that some negative opinion may spiral out of proportion. If something is not working right for the customer, whatever it be, it has to be stopped right away. Customer trust is the ultimate goal, and nothing can compromise it.

If some customer issue reaches Jeff, and he forwards it to the respective stakeholder with a question mark, then the executive is expected to drop everything and work to explain and correct the issue immediately.

As with everything else deliberately designed by Bezos, the Amazon Management System is also a self-reinforcing flywheel.

Further Amazon Management System uses 14 principles to define the corporate culture. They are

1. Customer Obsession

2. Ownership

3. Invent and Simplify

4. Are Right, A Lot

5. Learn and Be Curious

6. Hire and Develop the Best

7. Insist on the Highest Standards

8. Think Big

9. Bias for Action

10. Frugality

11. Earn Trust

12. Dive Deep

13. Have Backbone; Disagree and Commit

14. Deliver Results

Scroll to Top