2016 Letter to Shareholders

“Jeff, what does Day 2 look like?”

[ … ] Not shown: The question came from an all-hands; the Day 1 building.

“Day 2 is stasis.¹ (Editorial note 1)Editorial · 1Known at the timeThe threat named here is success, not crisis. Bezos describes a decline slow enough to “harvest” for decades, so nothing in the results warns of it. That makes Day 2 a matter of habits inside the company rather than numbers, and the four defences that follow are all habits. The cost is that the letter gives no test for when a company has crossed into Day 2.SourceAmazonJeff Bezos' 2016 Letter to Amazon Shareholders17 Apr 2017 Followed by irrelevance. Followed by excruciating, painful decline. Followed by death. And that is why it is always Day 1.”

To be sure, this kind of decline would happen in extreme slow motion. An established company might harvest Day 2 for decades, but the final result would still come.

[ … ] Not shown: The question of how to fend off Day 2.

I don’t know the whole answer, but I may know bits of it. Here’s a starter pack of essentials for Day 1 defense: customer obsession, a skeptical view of proxies, the eager adoption of external trends, and high-velocity decision making.

[ … ] Not shown: Customer obsession, with Prime as the example.

A common example is process as proxy. Good process serves you so you can serve customers. But if you’re not watchful, the process can become the thing.² (Editorial note 2)Editorial · 2Known at the timeBezos does not argue against process; he argues against a process that nobody owns. The diagnostic he offers is a question, “do we own the process or does the process own us?”, not a rule. That suits a company that runs on documented mechanisms, but a question is easy to ask and hard to enforce, which leaves the defence to each leader’s judgment.SourceAmazonJeff Bezos' 2016 Letter to Amazon Shareholders17 Apr 2017

[ … ] Not shown: Leaders who defend outcomes by process; surveys as proxies.

The outside world can push you into Day 2 if you won’t or can’t embrace powerful trends quickly. If you fight them, you’re probably fighting the future. Embrace them and you have a tailwind.

These big trends are not that hard to spot (they get talked and written about a lot), but they can be strangely hard for large organizations to embrace. We’re in the middle of an obvious one right now: machine learning and artificial intelligence.³ (Fact note 3)Fact · 3With hindsightAndy Jassy’s 2022 letter, published on 13 April 2023, returned to the trend: “a newer form of machine learning, called Generative AI, has burst onto the scene”, and Amazon would “continue to invest substantially in these models”. In June 2025 he told employees that AI would, “in the next few years”, “reduce our total corporate workforce”.SourcesAmazonAmazon CEO Andy Jassy’s 2022 Letter to Shareholders13 Apr 2023AmazonUpdate from Amazon CEO Andy Jassy on Generative AI17 Jun 2025

[ … ] Not shown: Amazon’s machine-learning work: drones, Amazon Go, Alexa, forecasting, search.

Inside AWS, we’re excited to lower the costs and barriers to machine learning and AI so organizations of all sizes can take advantage of these advanced techniques.⁴ (Fact note 4)Fact · 4With hindsightOn 29 November 2017 AWS launched Amazon SageMaker, a service to “build, train, and deploy machine learning models at any scale”. On 28 September 2023 it made Amazon Bedrock, a service for building generative AI applications on other companies’ and Amazon’s models, generally available. Jassy’s 2023 letter says Bedrock had “tens of thousands of active customers after just a few months”.SourcesAmazon Web ServicesIntroducing Amazon SageMaker29 Nov 2017Amazon Web ServicesAmazon Bedrock is now generally available28 Sep 2023AmazonAmazon CEO Andy Jassy’s 2023 Letter to Shareholders11 Apr 2024

[ … ] Not shown: AWS machine-learning services then on offer.

Day 2 companies make high-quality decisions, but they make high-quality decisions slowly. To keep the energy and dynamism of Day 1, you have to somehow make high-quality, high-velocity decisions. Easy for start-ups and very challenging for large organizations.

[ … ] Not shown: The senior team’s commitment to decision speed.

First, never use a one-size-fits-all decision-making process. Many decisions are reversible, two-way doors. Those decisions can use a light-weight process. For those, so what if you’re wrong? I wrote about this in more detail in last year’s letter.⁵ (Fact note 5)Fact · 5Known at the timeThe 2015 letter splits decisions into “one-way doors”, to be made “methodically, carefully, slowly”, and “two-way doors”, which “can and should be made quickly”. It warns that as organisations grow they tend “to use the heavy-weight Type 1 decision-making process on most decisions”, with “slowness” as one result.SourceAmazon investor relations2015 Letter to ShareholdersApr 2016

Second, most decisions should probably be made with somewhere around 70% of the information you wish you had.⁶ (Editorial note 6)Editorial · 6Known at the timeThe 70% is a rule of thumb, not a measurement: nobody knows what share of the information they lack. Its real job is to move the default from waiting to acting. It rests on the next claim, that being wrong is cheap if you correct fast, which holds for two-way doors and not for the one-way decisions the letter set aside a paragraph earlier.SourceAmazonJeff Bezos' 2016 Letter to Amazon Shareholders17 Apr 2017 If you wait for 90%, in most cases, you’re probably being slow.

[ … ] Not shown: Course correction makes being wrong cheaper than being slow.

Third, use the phrase “disagree and commit.” This phrase will save a lot of time.

This isn’t one way. If you’re the boss, you should do this too. I disagree and commit all the time.⁷ (Editorial note 7)Editorial · 7Known at the timeThe useful move is aiming the phrase at the most senior person in the room. “Disagree and commit” is usually read as a way to get dissenters to fall in line; here the CEO uses it to let a team overrule him on an Amazon Studios original. It only works if the disagreement is stated in full first, which Bezos insists on, and if the boss’s commitment is sincere, which no one outside can check.SourceAmazonJeff Bezos' 2016 Letter to Amazon Shareholders17 Apr 2017

[ … ] Not shown: Bezos’ example: an Amazon Studios original he doubted.

Without escalation, the default dispute resolution mechanism for this scenario is exhaustion. Whoever has more stamina carries the decision.

I’ve seen many examples of sincere misalignment at Amazon over the years. When we decided to invite third party sellers to compete directly against us on our own product detail pages – that was a big one.⁸ (Fact note 8)Fact · 8With hindsightBezos’ 2018 letter reports that independent sellers’ share of physical gross merchandise sales on Amazon grew “from 3% of the total to 58%” between 1999 and 2018, and says: “Third-party sellers are kicking our first party butt. Badly.”SourceAmazon investor relations2018 Letter to ShareholdersApr 2019

[ … ] Not shown: Quick escalation over wearing people down; four closing questions.

We can have the scope and capabilities of a large company and the spirit and heart of a small one.⁹ (Fact note 9)Fact · 9With hindsightIn September 2024 Jassy wrote that Amazon wanted to “operate like the world’s largest startup”, and that added layers had created “pre-meetings for the pre-meetings for the decision meetings”. He also set up a “Bureaucracy Mailbox” for employees to report unnecessary process.SourceAmazonUpdate from Amazon CEO Andy Jassy on return-to-office plans and manager team ratio16 Sep 2024 But we have to choose it.

[ … ] Not shown: Thanks, and the reprinted 1997 letter.