---
title: "Corporate Politics in Tech: Everything Nobody Tells You"
author: "Ethan Evans (former Amazon VP) · Ryan Peterman"
description: "I've seen things that make me want to close my eyes. You want to talk about the dark side, and that's okay. I hope no one would do this, but I know people do. That's my opening gambit for “I..."
source: "https://adhenawer.net/posts/original/corporate-politics-tech-everything-nobody-tells.html"
lang: en
---

TL;DR · key points

1. I've seen things that make me want to close my eyes.
 2. This is Ethan Evans, former Amazon VP, back for a second time because I knew he had a lot to share about corporate politics.
 3. I'm comfortable with the fact that I'm not a perfect person.

## OPENING THE DARK SIDE [0:00](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=0)

I've seen things that make me want to close my eyes. You want to talk about the dark side, and that's okay. I hope no one would do this, but I know people do. That's my opening gambit for “I want to take it over.” After this conversation, I'm a little more cynical.

This is Ethan Evans, former Amazon VP, back for a second time because I knew he had a lot to share about corporate politics. He shared what to do if you have a bad manager. He said, “You could consider this as mutiny.” He said he'd be transparent, so I really put him to the test. When you said we would just go deep on transparency, you were not kidding. Now there are some engineers watching this who are thinking: “I hate you. I hate the politics.” Do yourself a favor and get good at this. Here's the full episode.

I'm comfortable with the fact that I'm not a perfect person. Sometimes some of the stuff I did, I wish I had done differently. So I can share what I did and what I saw others do because I'm comfortable with that. And frankly, I'm retired. So I'm immune. There's no boss over me who's going to say, “You told Ryan what?” I'm free from that.

## EMPIRE BUILDING: WHY COUNTING HEADS RULES [2:37](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=157)

Empire building is a bad thing because it focuses more on the raw number of people than on what you're doing or the impact you're having. Why does it exist? Because it's rewarded. And it's rewarded because counting people is the easiest thing to do. When leaders decide who's ready for promotion, how much impact you had is subjective. One person thinks the project was hard, another thinks it was easy. But no one can debate that Ethan has forty-two people and Ryan has seventeen. Ethan is the bigger leader — that's the conclusion they're making.

Some companies have written thresholds: to be eligible for this promotion, you have to have this many people. Of course, if you give high performers a goal, they're going to figure out a way to hit it. And if that requires claiming they need people or seeking to take over other groups, they're ambitious. They'll find a way to rationalize it to themselves.

At Amazon, those thresholds did not exist while I was there. They existed in whisper numbers, but they weren't written down. After I left, people I was coaching told me the number to become a director in some orgs was eighty people, in others ninety. I became a director in the early days with twenty-two people, because it was different times and the company was smaller. The Amazon leadership principles say there's no bonus for additional headcount. But now that it's written down that I can't get to this very well-paying level without ninety people, there damn sure is a bonus. There's like a several hundred thousand a year bonus, quite literally, for accumulating ninety people.

I've talked to leaders whose boss, trying to get them across the threshold, said, “We're going to put these people under you for six months, check the box, and give them back to the person we took them from.” That's the game they have to play if they want to reward someone on their team. It's insanity — we're moving people around and changing their lives just to hit a number.

## THE HIDDEN CHESS OF REORGS [13:10](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=790)

The triggering condition for a reorg is usually a change in the business. Something has changed and the organization isn't set up to address it — a new market, an acquisition, a new project. The second reason is somebody else quit or got moved, so reorgs come due to cascades.

All managers know that too many reorgs cost you people, so you don't want to do them too often. What that means is you wait as long as you can, and then when you need to do one, you're asking: how do we get everything we want? If there's someone we don't like, how do we give them less so they'll move on? If there's a high performer, how do we set them up for success? How do we retain key people? Hopefully the start of the reorg is “how do we make the business better?” But as soon as we figure that out, we're trying to figure out how much else we can throw on the deck as the ship is pulling away.

There are usually multiple people involved. Sometimes you have to get them on board with your second motive: “Yeah, I want to help this person get promoted as a part of this reorg. Are you supportive of that?” Sometimes you have to tell a story like, “No, I'm not doing that, it's really this reason.” You're a storyteller, coming up with your narrative about why this is the best possible reorg — rewarding the deserving people and accomplishing the business goals. If anybody does suffer, you've made sure it's people who aren't the highest performers.

When I think about reorgs I've done, they're very much about retention. People don't like change. Who is it I absolutely need to keep? And who is it that if they leave, I'll either be okay or even maybe happy? I can't think of a time where I put someone in a role thinking “by putting them there, they'll quit.” But I can absolutely think of the reverse — people I really wanted to keep, using the reorg to give them something juicier.

There's a hidden chess going on. You're aware of multiple things, and aware of the messenger: this is what this person wants to hear, and here are the things I'm trying to achieve. What pieces can I sell to make the reorg happen? That turns off a lot of engineers. They say, “I don't want to manage all these things, I just want to say it as it is.” But you can certainly work around it — you can do a reorg purely on business goals without any concern for people, but that's probably going to cost you people.

## THE SQUEAKY WHEEL GETS THE GREASE [16:51](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=1011)

Sometimes two people perform roughly equally, but one of them has made clear they're thinking about leaving. A lot of managers will think: Ryan's this great guy, always sticks with us. He'll wait another six months, and I can use those six months to save this other guy. It's not that I mean to screw you. I'm focused on saving this other person and I know you'll put up with it. So you can end up behind just because you're such a nice guy.

I often tell my coaching clients: “You are a very nice person.” And what they don't know is I'm then going to say, “and that is not a compliment in this context.” I don't think you need to be a jerk, but pushy people get more — up to a point. They can push too far and get less because they've antagonized. But if you never ask for anything, you'll never get anything.

In a reorg, what's inevitable is I have some people who've been more vocal about what they want than others. Perhaps I'm arranging the reorg knowing what some people want, assuming the quiet ones will be fine. That's the benign version. The less benign is I know some are pushy and some aren't, so I'm serving the pushy people because then I get to keep my whole team.

## "MY CAREER IS VERY IMPORTANT TO ME" [21:13](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=1273)

I wanted to tell my manager Neil that I really expected to be promoted, but diplomatically. I was a senior manager at Amazon. I went to him and said, “Neil, I need to understand how important my career is to Amazon, because my career is very important to me.”

That's a statement no one can argue with. No one can say your career shouldn't be important to you. “My career is very important to me. I need to understand how important it is to Amazon, because if it's not as important to Amazon as it is to me, I need to think about that.”

“I need to think about that” isn't a direct threat, but everyone understands I'm serving notice. If I said, “If it's not important to you as it is to me, I might quit” — that's a threat. All managers will say, “You can't extort me, that isn't how it works.” So I came up with words that clearly conveyed the message but in a way you can't argue with. And he promoted me. He pushed the promotion through.

There's always a way to say things in that middle ground that makes clear what you want and that you're willing to consider your other options, without being a threatening jerk. You have to tune your message to the workplace. That message would be offensive in some workplaces, and it was totally not offensive at Amazon because there were people pounding the table. Find a place where you look like a wonderful sweet person in the mix.

## THE ART OF MAKING DEALS [25:24](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=1524)

The great thing about really demanding people is if they're bothering you to the point where they're a net negative, you just tell them no — they will quit. They are so driven that if you deny them, your problem will go away shortly. They'll even leave for a lesser job.

But I'm a huge believer in deal making. I made these deals with my own leadership too. The deal essentially is: “Okay, you want this thing. I will help you get it, but you're a high performer. You're going to show incredible performance on the things I need, and then I will deliver.”

My manager gave me a requirement I didn't like. He said, “Your team is too expensive. You need to open a low-cost offshore dev center.” I didn't want to do any of that. I found a senior manager I knew who wanted to get to director. He was Indian, living in the US. I said, “Would you go back to India and start this dev center? If you can build me a dev center in India, I think we can use that as a platform to get you to director.” That's exactly what happened. He moved home to India, took his son home to learn Hindi and see his grandparents. He moved for three years, and I was able to create a story about how this was clearly a director-level accomplishment. He was promoted to director. I made that deal years before it happened.

There's also the forward-looking slate. HR in big companies tracks, by halves, who's in six months, twelve months, eighteen months, maybe as far as twenty-four. There was a sense of whose turn it was. I understand at Google today there's a strict quota — this quarter, two people can get from L8 to L9. Someone I coach said, “I think I'm number four, I'm not going to make it.” That can be toxic, but people understand the idea of a line. Things get more quota-based when positions are rare and expensive. Principal and distinguished engineers cost a million dollars apiece. We don't just hand them out like popcorn.

## PIRATES IN YOUR OWN ORGANIZATION [33:05](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=1985)

This has become more common as growth has slowed. When Amazon was growing very quickly, you saw less of this — it's easier to go get new people than to fight with someone. Now that companies are flat, there's a lot more of what I'd call cannibalism. Since I can't get new, I'm going to eat what's here.

You can't allow people on your team to go rogue and be pirates — capture somebody else's ship, kill the captain, sail away. Because of what it does to your org: you're no longer in charge, they are. They're raiding and pillaging and you seem clueless.

What holds back the worst behavior is people realize: if I get caught being too obviously self-serving, it's going to burn me here. It's like any coup attempt. If you win, you get to be president. If you fail, you get hung or shot.

Instead of saying “give me this,” tell your manager: “I want what's best for the org. If you have a need, I'll be there for you.” That sounds selfless and org-centric, even though the underlying message is “I don't really care which team's available, I just want more scope.” There's a polite fiction here that can be true: I understand that my best path up is to do things that are valuable to you and to the organization.

When a reorg comes, they're like, “Oh right — Ryan said he'd stay. Perfect. Ryan just became most qualified, not because he's best, but because he's most reliable.” And that is best, in my situation.

## SAYING YES: BUILDING THE GARBAGE-CAN EMPIRE [37:59](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=2279)

When I was a director working toward VP, I was put under a VP who had no other engineering leaders, but they kept giving him more engineering stuff. I was the first, I built trust with him, and he automatically came to me with everything. “They handed me this, can you take it?” At one point I was managing eight different functions.

I led the technical team for reverse logistics — when people traded in video game cartridges, we had to grade them and sell them used. I was selling new video games as software downloads. I was starting to build Amazon's app store. I owned our B2B business, where Amazon was trying to displace Office Depot and ship pallets of paper. These all ended up under my boss. I had this random mixture, but my boss loved it, and over time that totaled two hundred people. That came from saying yes to everything.

Someone I really respect, now a senior VP at Walmart, was taught by a senior VP at Amazon: “The simple way to get promoted is, solve problems for your boss.” If you need to get it down to five words: solve problems for your boss. If you're solving their problems, they will value you. Most people are not helping — they're punching the clock or fighting for their own things. Even a bad boss will often take care of someone who's helping them. Even if I'm evil and only care about myself, who do I want to keep? The ones that are helping me.

## THE INVENTOR PATH: AMAZON T-SHIRTS [42:02](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=2522)

I also recommend being the inventor. Amazon now sells over a billion dollars a year of custom-printed funny t-shirts. I helped start that business. My manager said, “Hey, you run Amazon's app store. What does that have to do with t-shirt printing?” I said, “Well, I think there's a big business here. I need ten people and I have eight hundred. If you're telling me I can't spend basically 1% of my resources on this thing I believe in, we have another discussion, because you're micromanaging me.”

He said, “Fine, I disagree, but go do your thing.” It turned into a big business. I'm simplifying a ton — there were years of work in that phrase. But I was always thinking about what else I could do with my own resources or lightweight approval. I think there are multiple paths to the top. I tried them all. Be the helpful person, but also be the inventor.

## THE GARBAGE CAN AND THE WINDOW SEAT [45:52](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=2752)

If I wanted to get rid of a pushy, overly ambitious person, I'd just give them “other.” Six weeks later, that's a vacancy. When I first joined Amazon, there were a few groups: one person would own what became Prime Video, another Prime Music, another the Kindle. Then there was one more leader who got given database maintenance, quality assurance, and a couple small things. Even he joked, “Well, my group is 'other.' It's the garbage can.” Our VP said, “Don't call it that. What about the people in it?” But as far as I know, that guy is still at Amazon twenty years later because he was one of those calm, nice guys willing to do what was needed.

This isn't an Amazon thing. In Japanese culture, there's a concept called the “window seat.” You want to be in the center of the building, close to the CEO and close to the action. Being on a window means you're not that important. What Americans would see as a great corner office is, in Japan, a signal that you've been moved as far from what's happening as possible — exiled to the land of irrelevance, where you don't even have co-workers on one side. That would signal your career here is over.

## MUTINY: NEVER GO ALONE [52:50](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=3170)

If you have a low-performing manager, the natural thought is to go to their manager — your skip. You could consider this as mutiny.

When people come to me about a weakness in my employee, there's a subconscious process. I have two choices: I can believe you're overly sensitive and high maintenance — in which case you're the problem, and if you quit, the manager has to backfill. Or I can agree with you, and now I have three problems: I have to decide what to do with my manager, I have to hire and train somebody else, and while they're gone, I have to do all their work myself. You can see why, even subconsciously, I have reasons not to listen.

Never mutiny alone. If you really want to say “this manager is bad,” take two or three people, or send them in sequence. I had a leader who was treating the women on his team badly. It took a long time for different rumors to bubble up. Had any one of them come to me alone, I probably wouldn't have listened. But when several reports came up, I realized I'd been missing something. Even in Zelda — it's dangerous to go alone. Take this.

## GETTING OUT FROM UNDER A BAD BOSS [57:22](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=3442)

There's a better approach than mutiny. Don't even bring up the manager. Say, “Hey, I was looking at this other role and I think I could do so much more for you over there because of A, B, and C.” Try that first. If it works, you never have to bring it up.

You can play chess, not checkers — several moves in advance. Once I did this with a VP. I was a VP reporting to him. I had gotten a job offer from an SVP in a different organization. I went to him and said, “I have an offer from this SVP. Now, you know that if I give my word to that guy, I have to go. That's what you would do. If I give my word, it's over. You've told me no on this other role several times. What do you want me to do?”

He understood he would never break his word to an SVP. The moment I gave my word, it was done. After weeks of resistance, I was in the new role the next morning. Once he got clarity that the choice wasn't “you stay in my role or go to the role I don't want you in,” but rather “I keep you in my team or I lose you,” everything changed.

I was careful not to make it threatening. I didn't say “unless you give me what I want.” I said, “once I give my word, I can't go back on that.” I tap into the manager feeling like, “Okay, I don't like it, but I get it.”

## THE POLITE FICTION [1:02:13](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=3733)

A mistake people make is not thinking through their wording before they're in the room. Most of these conversations are not spur of the moment — they're on the calendar, or you know they're coming. Spend the time to have that key phrase ready so you can deliver it with true sincerity.

When I said “once I give my word, I can't go back on that,” I looked like I meant it. In no way was I lying, but I delivered it like I'm an angel just fallen from heaven. Poor innocent victim. Both are true.

When I put the screws to that other boss long ago — “I'm gonna have to consider if my career is not as important to Amazon as it is to me” — I said that with a warm smile: “Of course I'm gonna have to think about that.” I looked like an anime kitten with the giant eyes. Body language matters. Same lines, same words — if I deliver it angrily, you're thinking “screw you.” If I deliver it warmly, it's totally different.

Some engineers hate this. I hear you, but this is human nature. You're dealing with humans, with their emotions, defensiveness, and stress. Do yourself a favor and get good at this. It's all through your life. Any negotiation, any deal making.

## INTERPERSONALLY WARM, PROFESSIONALLY FIRM [1:05:05](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=3905)

I call it being interpersonally warm: friendly, smiling, not getting agitated, while being professionally firm — saying what I need about what I really want, but not bringing anger or “you versus me” into it.

A great phrase is “help.” “How can you help me? I want to get to the next level.” Everyone loves to think of themselves as helpful. That's different from “What are you going to do for me?”

I read a book called “Leadership and Self-Deception” by the Arbinger Institute. What it taught me is: don't see the other person as an obstacle to what you want. See them as a human being trying to do the best they can in their situation. When I stop seeing my boss as “this person controls my pay, I have to yank it out of them,” and instead as a human trying to get by, I can say, “Look, here's what I need. What do you need so we can make this work?”

We objectify others. As an engineer, you do this all the time: “I just need them to do this thing. How can I get them to do this thing?” That's making them an object. I'm treating them like an API — make the right call, get the right result. Instead: “Look, I know you have a busy roadmap, I understand you have pressures, but I do need this. So what can we do together?”

Nothing works 100% of the time. I was debating politics with my college roommate and decided I'd use all my influence skills. I failed. Nobody left angry, but I got nowhere. Sometimes their motives just can't be aligned with yours. These techniques will get you way more success — but they're not a magic recipe.

## INFLUENCE, POLITICS, AND MOTIVE [1:12:22](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=4342)

People ask me: what's the difference between influence and politics, between influence and manipulation? The skill sets are the same. It's about motive.

I teach executive presence. Darth Vader had it in spades. So did Palpatine. These powerful, commanding figures. The skills of being influential can be used for good or evil. When I meet a skilled operator, I'm trying to figure out: US special forces or Russian special forces? Who am I dealing with?

There are three cases. People whose motive is to do good — easy, great. People completely out for themselves — Sith versus Jedi, lightsaber up. And the middle ground: this person isn't altruistic, but in this case they kind of want what I want, or I can give them enough wins to negotiate. We're never going to be best friends. I don't think you're the best person. But in this limited scope, we can ship something together.

If you could separate from an unethical person, great. But at work we don't always get to choose our teammates. Sometimes you need a team run by a terrible or self-serving person. Learn to work with them. With slippery people: they love to be slippery in the dark. Bring allies so it looks like “everyone thinks this.” They usually can't afford to have everyone know they're unethical. There's a phrase from politics: “plausibly deniable.” Put them in a situation where it's easier to go along with you than fight you. “I'd rather stab Ethan in the back, but there are a lot of witnesses.”

## BE A PORCUPINE: ORG VS ORG [1:20:30](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=4830)

A very smart engineering leader once said: “Respect between two engineering teams is inversely proportional to their distance.” The further apart they are, the more they belittle each other because they can't see the work. I've heard Amazon S3 described as “just a big disc drive.” Of course, at some level it is — but that's so dismissive. The closer you get, the more you see the edge cases: “Oh, actually that is complicated.”

When one org wants to take over another, very few humans wake up saying, “I'm going to be evil today.” They want something and construct a line of logic about why it's actually great.

How do you protect your org? First, the skunk trick: make yourself unattractive. List all the unsexy things you maintain — “We actually run the Yugoslavian tax engine. Lots of tickets there. Are you ready to take that on?” Second, be thorny. Make yourself the least attractive target. Counterattack warmly: “It's so good you're thinking about merging our orgs. Your org is going to be so great as part of our org, and I promise I'll take care of you.” Now they're thinking: “Maybe I messed with the bull and got the horns.”

Where you're vulnerable is if your work is behind, not making money, performing poorly. On a strong platform of high performance, you have arguments: “Don't distract us. Reorgs cause a six-month hit in productivity. We're killing it on profitability. We shouldn't even talk about this for six months.”

People love to think facts make decisions. But people come to an emotional conclusion and then rationalize it. If I'm a better storyteller, I give you a narrative that feels emotionally right. I help you wrap your own story around it. Once it's your story, it's very hard to shake. Be the better storyteller.

## ATTACKER TACTICS AND THE "I WANT TO HELP" GAMBIT [1:27:20](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=5240)

What are the attacker tactics? The best attack — which isn't even necessarily an attack — is a real story of how merging creates more value. “This group puts out one release a year, I can get it to two. Here's how.”

Slightly less good arguments are about efficiency: I can save costs, consolidate, fewer direct reports. These aren't false, but they're less about business growth and more about making things smoother or cheaper.

I hope no one would do this, but I know people do: go highlight flaws in the other org. Show they have turnover, high cost, a recent outage. You're basically trying to say “it's a dumpster fire, let me put it out.” But what if it is a dumpster fire? Then the way to do it is not to say “I should take over.” Go to the common boss and say, “Hey, I can't help but notice you're having trouble here. Is there anything I can do to help?” That's my opening gambit for “I want to take it over.” I go in and start helping. Later they're saying, “Wow, it's so much better under Ethan's leadership.” Then I can casually mention, “If it would really help you, we could just consolidate. Now that I'm here, I see all these efficiencies.”

## DEALING WITH TOXIC PEOPLE IN OTHER ORGS [1:30:26](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=5426)

When there's a person in another org who is well-liked there but is a major problem for your org: one of my managers called it a “meat shield.” He would put someone in a position and say, “Your job is to keep them off of us. It's going to be a terrible job, but I'll reward you later.” Go be a meat shield until this project is over.

The most important thing if you want change is understand why that person is valuable to their org. What they're telling themselves is, “Well, they have sharp elbows, but they get so much done that's valuable. It's okay.”

Go to that leadership with options: “Look, we're willing to give you this value without all this pain. Is there a better way?” Or: “Your guy is driving us nuts. That's going to change. Either you give us a new interface, or you move them away from us, or we're going to stop cooperating. Everything will be late. Everything will be your fault. Everything will require six document reviews. I'll spend the rest of the day asking ChatGPT what the most policy checkpoints I could put on our next release would be.”

My external narrative would be: “Dear peer, this person is really making things difficult. I know they're a valuable member of your team, but our team is struggling with them. Option one, you find a new interface person. Option two, we'll have to distance ourselves, and unfortunately that creates friction between our orgs. That takes both our time, and I don't want to take your time, but I have to protect my people. Some are ready to leave. What would you do if people on your team were ready to leave? I'm going to have to take steps that increase bureaucracy. I'm sure there's a better way.”

You didn't bluster. You didn't threaten. Just by circumstance: if you don't comply, your life will be harder. Most leaders at a certain level understand unspoken messages. They know I'm giving them a chance to collaborate but there's a stick up my sleeve.

## UMBRELLA MANAGERS VS FUNNEL MANAGERS [1:36:52](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=5812)

A lot of people don't know how to be socially skilled, so they can only say hard things once they let themselves get angry. “Ryan, you've missed three dates in a row.” I can only do that once I get internally angry. Some people need anger as fuel to be ready to fight back. You want a manager who can stand their ground because they believe in it and are willing to have confrontation without needing anger.

There's an image: all managers are either an umbrella or a funnel. When the poo comes down, they're either shielding their team or dropping it onto their team. The funnel manager is deciding: push back on my boss, which feels scary, or push back on you being upset about your fourth weekend of maintenance. You're an easier target because I have rank, I control your performance review.

You have to change the balance of safety. Help them feel safe pushing back on their boss, or help them feel unsafe asking for your fourth weekend. “Ethan, this is a pattern — we keep getting asked to do weekend work. I would love to help you push back on that. I can give you data. But I have to be clear: my family is very important to me. I can't keep working weekends. What I don't want is for you to be in a position where you accept work I'm going to decline and then you have committed to something that won't happen. How can we plan to push back next time?”

The wrong time to start this is while the request is in hand. Do it Tuesday or Wednesday, after the weekend and before the next crisis. If you're not willing to have the manager say “If you don't like it, quit,” you can't say anything. You have to have enough confidence that if it comes to that, you're okay leaving. Good engineers are valuable. Teams that are turnover factories aren't effective. If you're at a place where you can't create value, the right answer is to get out. There's always more jobs. One of my beliefs is: to work less, you often have to work more first. To get to the place where you can coast, you have to do some ugly digging first.

## BACK CHANNELING [1:46:47](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=6407)

I think of back channeling in two ways. Manipulative back channeling is where I'm trying to do something I wouldn't be willing to say in public — undermining, painting the other leader negatively.

Positive back channeling is giving people a chance to ask questions and find out their concerns without bringing them out in public. Maybe I'm the CFO concerned about the budget but I don't want to put you on the spot. In a back channel, I can ask: “Here's my idea. Any concerns?” I'm giving you a chance to feel understood, consulted, listened to.

Back channeling doesn't just work because of information exchange. You made them feel important. That emotional buy-in — particularly if the other project leader doesn't do it — gives you a huge leg up. You're building trust. Even if the other proposal is a little better, remember part of my bet as executive is: who can I count on when the chips are down? Ryan came to me and asked my input. Totally, I'm funding that, even though the payoff is a little lower, because I can work with him.

The negative version is the backroom deal — the kickback. “You support this and I'll make sure you're taken care of.” I'm coming with a proposal that amounts to: you should do the wrong thing because we both benefit. Even though I'm a snake, I'm not going to bite you. We're snakes together.

Some people are up for that. Some people look at the world and say: I paid my dues, I got screwed over when I was weak, now that I'm strong it's my right to take all I didn't get. They always tell themselves a story. It's not “I'm screwing those people,” it's “I'm getting what I'm owed, this is only fair.” When you're dealing with someone vicious, realize their story to themselves is not that they're evil. If you can figure out their story, you have a better chance to redirect it.

## OUTSIZED INFLUENCE FROM IDEAS [1:52:38](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=6758)

I've seen soft power come from ideas. Amazon built a product called Cloud Drive, a storage system like Dropbox. That proposal came from an entry-level new college graduate engineer. He did two things. Amazon had an invite-only engineering conference. He fought his way in by making a proposal. Then he knew Jeff Bezos came to review an idea poster session. He emailed Jeff: “I have this idea, I'll be at the poster session. Could you stop by?” A very small ask of the CEO.

Jeff liked it and emailed his triple-skip — my boss. So there was Jeff, then a VP, then me, then a manager, then this new engineer. My VP didn't like the idea at all. But Jeff had said “look into it,” and the project got funded and built.

I worked with a slightly more senior person who helped put forward the proposal for Amazon Fire TV. He wasn't super senior, but he did the legwork and had the connections.

Influence can come from anywhere. It usually begins with a good story based on real fact. Told with enthusiasm, repeatedly, specifically — to people with the power to do it. “Wouldn't it be great if we had a cloud storage system? Wouldn't it be great if we had a low-cost living room device where people could watch Prime Video?” An understandable soundbite story everyone agreed with on the surface.

## WHY PEOPLE REALLY GET FIRED [1:55:41](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=6941)

There are a lot of rules about when you can and can't fire someone, so almost always the public story is performance. That's the one thing that's allowed.

The hidden reasons are usually incompatibility. The main one I see is style differences between the person and their leader. One is detail-oriented, one is high level. With engineers, often one wants to be tech-oriented — what's the best technology — and the other is business-outcome-oriented — what makes the money, I don't care if it's a hack. It becomes a religious debate. Over time, small things become labels: “that person is technically clueless,” “that person's only about money.” Once it's personal, you have a lens. So many things are ambiguous, and you'll interpret everything they do as bad. That's where people get fired, but it's always labeled as performance.

One of the most blatant harassment cases I saw was an executive hired at chief level who immediately started harassing women. So blatant that when we asked one of the HR representatives on his team if she'd gotten complaints, she started crying because he had harassed her too. The gall to harass a woman in human resources. He was terminated right away.

How did he get hired? We didn't have legal proof. We had many people saying it happened, we believed them, but never took it to court. If I name him and say “don't ever hire Fred Smith, he harasses women,” Fred can sue me — I have no legal proof. So when someone calls asking about Fred, I refer them to HR, and HR says “Fred was employed from this date to that date.” That's how these people keep getting hired. I looked him up on LinkedIn — the fact he ever worked at Amazon isn't on his resume. He just deleted it. Moved on.

## FIRING MANAGERS IS HARDER [2:05:57](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=7557)

It is harder to fire managers. Three reasons. One: they're more experienced. They don't have obvious bad performance — the first person I let go stopped coming to work. Managers don't make mistakes at that level. Two: a manager's work is hard to measure. Did they hire? Did features ship on time? Bug count? It's a “he said, she said” problem. Three: managers are part of the system. They know what termination looks like — they get spidey sense and start positioning themselves to be hard to get rid of.

My ideal way is to surface the idea socially. “I don't think we've gotten off to the best start here, and I'm wondering how you feel about it. Would you like my help finding some other option?” Very nebulous. Smart managers get the message: I'm your boss. I'm not happy. I'm offering you a door — time to find a job, maybe a severance package. Do you want to do this the easy way?

They won't say “Of course, you're right, I'm failing.” A smooth manager will say, “I'm so sorry you're not satisfied. Of course I'm doing my utmost” — meaning, I can make this hard for you, you're going to pay for my exit. “I want you to be happy with my work. If you're not, I'd like to discuss what changes we could make” — meaning, I'm open.

I never used the word “fired” or “performance,” because that creates something legally called constructive termination. If I've said “we might have to fire you, do you want to quit first?” that's a legal problem — anything I do from that point is seen as me making up reasons. So I avoided it. They equally avoided acknowledging any problem. “I'm sorry you feel that way.” That's your problem, you feel weird things. Both sides know exactly what's happening.

The other path is classic performance management — document goals, document they missed them. Almost all managers will quit before that completes. They know “this is going to take ninety days, I have ninety days to find a new job.”

## YOUR LEVERAGE WHEN BEING FIRED [2:12:31](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=7951)

I'm a little more cynical. They're paying severance for something. What's the leverage?

First: this process takes a lot of manager time. You have to document, follow process, file paperwork. HR audits. The first offer I make: “I'll save you all that time if you give me money to look for another job.”

Second: the story I tell the team. “I can leave talking about what a terrible experience this was and what a monster you are. Or I can leave saying this wasn't a fit for me, I'm excited to be moving to Acme Inc, and I can't speak highly enough about how Ryan treated me through this. Which story do you want me telling my co-workers?”

You deliver it as polite fiction. “Ryan, so far you've been nothing but supportive. If I were to move on someday, I would of course explain how supportive you've been. But if we're discussing my moving on, I would hope you'd want to support my transition financially so that I can tell that story.”

Buy time. If they're giving you performance feedback, say: “Ryan, I appreciate you're giving me this feedback. I know we need to have this conversation. Honestly, this moment I can't have it productively. Can we have it tomorrow morning?” Most managers won't say no, because they're already on edge. If they say no — now they have a beef with HR. “He was so mean. I asked for space. He said no, I've got to ridicule you right now.”

Your last piece of leverage: your manager doesn't want to feel bad about themselves. They want to feel good. If they can spend a little of the company's money — which isn't really their money — so they can look in the mirror and say, “It didn't work out with Ryan, but I took care of him,” they want to do that. Most managers hate firing. They avoid it. One reason people don't survive firings is by the time the manager decides, they've told themselves they're out of options. Getting back off the brink is very unlikely.

People come to me: “Ethan, I've just been put on a performance improvement plan. Help me survive it.” I tell them: “Honestly, no. You're most likely going to be fired.” By the time they get around to a formal plan, they've tried everything else. They just want you gone as quietly as possible. People write back three months later: “You were right. I was fired.”

A softer phrasing for you: “This is such a hard message, and I have so many friends on this team. I don't want them to think badly of me. I don't want them to think this is an unfair environment. Can you work with me so this is a graceful exit?” Later: “For it to feel good, I need enough security to get to my next thing. I know you want that. I know you don't want anyone to suffer.”

## GETTING UNSTUCK AT STAFF ENGINEER [2:24:37](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=8677)

What gets people stuck is they're not learning the new skill of the next level. They're doing the current skill harder. What gets you promoted early is hard effective work — delivery, shipping, building. But as you get further, you need different skills: delegate, communicate, let go of your hard skills. If you micromanage — “write it this way, build it that way” — you'll drive away your good people.

There's a classic book by Marshall Goldsmith: “What Got You Here Won't Get You There.” It shows a ladder with a gap. Talk to people a level or two above. Ask them: “What did you have to let go of? What did you have to learn? What do you see holding me back?”

Getting advice is better than getting feedback. If I ask you for feedback, I'm asking you to do something hard — criticize me. If I ask for advice, everyone has advice to share. They'll give you the same information more positively.

I see so many people working brutally hard but ineffectively.

## HOW TO WIN AGAINST AN EQUAL PEER [2:27:36](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=8856)

Assume you're both equally qualified. You're not going to win on skill. So: relationship and trust.

Trust is: do I have to look over my shoulder wondering if you're doing things that won't make me look bad? Will we both be successful? As a senior leader, I don't have time to micromanage. I need to give you something and assume it will get handled.

Trust is earned more in bad times than good. My favorite analogy is violent: two people in a war together. When the bullets were flying, you came through — I know you'll come through the rest of the time. At work, that's a crisis, a big outage, a lost account. If you jump on that and fix it, even if it costs you — you cancel a vacation (not that I recommend doing that a lot) — I will remember: when the chips were down, when we lost the Acme account, Ryan flew back from Hawaii early and made it. That's trust.

Second: do I like you? If I have two equally qualified people and I'm going to spend more time working with one, why would I not choose the person I enjoy? Some engineers hate this idea — if they were dead, they'd be spinning in their graves. Look, if I have to spend eight hours a day with someone, I'm going to be drawn to the person who's fun. It doesn't mean kiss up to everyone. It does mean: be friendly, be pleasant, be that helpful smiling person.

Bezos put it: when you come into the room, do you suck all the oxygen out of it? Or are people like, “Oh, Ryan's here — this just got better”? Simple rubric: do people want you in the room? If they kind of don't, you're in trouble. If they kind of do, you're doing great.

I know a lot of engineers get pissed off by this. I'm not here because you'll like me — I'm here to get you where you want to go. I'm not here to apologize for the way the world works. The way the world works sucks, but it does work that way.

## GETTING CHAMPIONS FOR PROMO PACKETS [2:32:54](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=9174)

For senior promotions, you need champions or allies — people at that level or higher who'll vouch for you.

My favorite way: your boss on your side. You've got the team where your manager wants to get you promoted. I would email my peers or skips: “Ryan's going to be up for promotion in six months. Because your group is adjacent to ours, I'll need to ask you for feedback. So you've got six months to watch Ryan.” Number one, me asking doesn't feel self-serving — I'm talking about you. Number two, six months later when that VP wants to say “I only saw Ryan twice,” I can say, “I put you on notice.”

I got good at getting my people promoted, and that allowed me to attract more talent. A guy who didn't get promoted — I said, “Look, I'm good at this, I'll get it done in six months, but you've got to give me till the next cycle.” He said later: “That sucked, but I decided I'd give you six months.” And we got it done. My biggest star employee went from level-five engineer, to senior, to manager, to senior manager, to director in eight years. Three promotions. He caught up to me and went to found his own company.

If your manager isn't that on board: go yourself. “I'm up for promotion. I'm pretty sure my manager will ask you for feedback. What can we do over the next six months so you're in a natural position to give it?” You need to oversubscribe — if you need four champions, ask six or seven, because one will leave, one will reorg, one will say no.

VPs know this is how the system works. They need your VP's help with their own people. They're good at understanding “this is the grease that keeps the thing flowing.” They can write mild approving verbiage pretty quickly — maybe it looks like they cut and pasted and replaced “John” with “Ryan,” but they've written a million of these.

I've never seen an engineer get promoted over their manager's objection. I can't promote you if the manager doesn't agree — I'd just be creating a mess, telling the manager they're wrong. Engineers envision: “I'll do such incredible work that I'll be recognized even though my manager is a bozo.” It's not going to happen. You either get your manager on board or get out from under them.

## AVOIDING POLITICS: LEVERAGE AND GROWTH [2:44:02](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=9842)

Can you avoid politics? First: choose your leader. Find someone who shares your disdain for politics — maybe with a tech background — and wants to reward work. Second: go work on the hardest area where expertise will matter. Stand out.

Andy Jassy, before he was CEO of all Amazon, was CEO of AWS. He said principal engineers are normally expected to advise, mentor, do architecture — but “these few guys working on AWS, we just need them solving the hardest problems. I absolve them of all the other stuff.” For engineers who don't want to review architecture or give promo feedback, he was the perfect boss. Their work was so essential he couldn't spare them.

If you have a skill no one else has and that skill is valuable, you'll get exceptions. I remember Andy talking about a high-level employee. Some criticized him — “he's not working at an Amazon level.” Andy said, “Yeah, I think you're right, but he's in a role where we need him. He knows we need him. We're going to have to live with it.” Andy was one of the hardest managers I worked for, but in this case: “Good on him. He's got us. We're going to have to live with it.”

You have to have leverage. Live up to Cal Newport's “So Good They Can't Ignore You.”

Growth changes everything. It's not twenty-five people wanting one promotion, it's lots of opportunity. My extreme example is the post office — the only roles are “carry the mail” or “be the postmaster.” There's one postmaster. Until that person leaves or dies, you don't get that job. In fifteen years at Amazon, the company grew from 14,000 to 1.4 million people. A hundredfold. My first team was six people. Nine years later I had 800. That growth opens doors that in a normal company growing 5% or 10% a year will never happen. Find high growth.

## THE ONE THING HE WISHES HE KNEW EARLY [2:48:23](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=10103)

What would I tell early me? That it's not just being right. Not just being technically astute. There's all this other stuff going on — wake up to it. I was blind or deprioritized it. I thought, “That's management stuff.” Even as an individual contributor, understand: management stuff does control the world. I wish I had been more aware.

To get it down to a single sentence: take it seriously that your relationships are what actually control your progress.

As an engineer, you want to believe it's your technical expertise that controls your career. It's not only that. It's your people skills. IQ without EQ is a waste of IQ. Cal Newport wrote “So Good They Can't Ignore You” — but the irony is he's a guy with the social skills to teach at Princeton and write books. He has the human skills to present his technical accomplishments delicately and well.

Humanity doesn't change. Read any ancient text of philosophy or religion — humans have been mean to each other since the first ape clubbed the other ape. I'm sorry that it frustrates you, but you're going to have to do it.