Taro says that “promotion is about behavior, not output” - How does landed impact factor in here?

I'll introduce a few more terms as well :) To land a promotion, you need three things that are forward-looking:

• Trust
• Scope
• Skill

You need one thing that is backward-looking: impact. The impact is what gives credibility to your level of trust, scope, and skill which are needed for career advancement.

When you have sufficient scope and trust, your behavior starts to change, which is why we have that phrase “promotion is about behavior, not output”.

Relevant links:

• [Masterclass] How To Get Promoted Faster In Tech
• A Huge, Often Unused Way Your Manager Can Help With Promotion

I love this question! It's true that impact is what truly matters at the end of the day, especially for L5 (senior) promotions and beyond. However, the purpose behind that saying is to incentivize the right kind of growth for software engineers looking to get promoted.

One of my best managers at Meta put this beautifully as she was breaking down the E4 -> E5 promotion (mid-level to senior at Meta) for me. She told me to envision 2 different E4s looking for E5 promotion, both of whom are getting "Exceeds Expectations" ratings:

• E4 #1 (fake growth) - This engineer grinds super hard and writes a staggering amount of code. With 250 diffs landed per half, they are writing twice as much code as the average Meta E4. However, the projects they're coding for are relatively straightforward with not super high technical complexity. They are also never taking a leadership role on any of the projects, just having the tech lead break down the tasks and ingesting them from there. This engineer is simply taking what an average E4 is doing and doing far more of it. Top companies like Meta do not want engineers getting promoted in this way as the value added here isn't special and it certainly isn't sustainable. They will get rewarded for the extra productivity and impact via the higher rating, but they almost certainly won't get promoted. All of the impact from this engineer comes directly from themselves (singular), not empowering others around them (multiplicative).
• E4 #2 (true growth) - This E4 "only" has 125 diffs per half. However, they are tackling much harder projects than E4 #1 where putting out a working diff is much harder. They are also taking a leadership role on these projects, crafting milestones and talking with XFN stakeholders to disambiguate the problem. On top of that, they are leading and mentoring E3s to take on the more straightforward coding work. A lot of this is rocky (this is why they're not E5 yet) as some of the E3s push bad code due to poor instruction and this E4 misses some crucial system design edge cases, but you can see the deeper behavior blossoming. This kind of growth is encouraged by top companies like Meta as this engineer is adding a rarer type of value most other E4s can't add. It's also far more sustainable as this engineer is finding more multiplicative ways to add impact, uplifting other engineers instead of just trying to solo-carry.

This is the reasoning behind "Promotion is about behavior, not output". It's best for the long-term future of the company and your personal growth to be more like E4 #2 instead of E4 #1. Again, impact is the most important, but the idea is to treat impact as a lagging indicator. If your behavior is truly evolving, then the greater impact will naturally follow. If you focus only on impact, you're very likely to take the path of E4 #1 and brute force increased impact by pouring more time into the problem.

On top of that, being able to land impact is a behavior in and of itself.

In this world, there are closers and there are chokers:

• Closer - Stays cool under pressure, maintaining a steady hand. The stress of barreling towards release doesn't get to them, and they are able to ship a final result the team can be proud of.
• Choker - Crumbles under the pressure at the last minute and is unable to ship the final result. Launch falls apart.

Ferrying along a project pre-launch honestly isn't too hard. There isn't too much value in having an engineer who can hold a status update meeting here and there but everything they try to launch completely implodes the second they cross the finish line. Being a choker is a trademark sign of a weaker software engineer who isn't ready for promotion.

Launching is when there's true stakes. Tons of users descend on your new feature code and push it to the brink. Any mistakes you made during planning like forgetting to account for a security flaw or user flow edge case will be exposed. This is when everyone actually knows if the behavior you demonstrated while building the project was good enough. This is when people find out if you are the closer instead of the choker.

Even if a launch goes wrong, this is another opportunity to show deeper engineering behaviors. I have seen Staff+ engineers have launches that blow up. The difference between the Staff+ engineer and lesser engineers is that they are always able to quickly solve those unexpected issues, leading their team through the chaos to efficiently find a solution. Lesser engineers get overwhelmed by these issues and often run away (I have seen this from a lot of E4s).

This is another way to view promotion for very senior levels: You are able to take projects of higher and higher complexity, spanning across multiple teams and even entire orgs, and continue to close. Even if there's 50+ stakeholders and a VP breathing down your neck, you are able ship something of high-quality at the end of the day, on-time and with respectable quality. You never crack under the pressure and are a fearless leader who inspires the right actions amongst your team.

If you want to be more of the closer instead of the choker, there's a lot of good resources here: [Taro Top 10] Project Management

That makes a lot of sense - thank you both for the explanations!