A bit of a cynical take (on Hacker News no less) but after being in the industry for a while, my view is that the best definition of “level” is self-referential: it corresponds to the ability of a person to convince others that they are at that level.
I also have a definition of what I think level should ideally represent: the marginal contribution of a person’s influence on the outcome of a company, relative to the counterfactual situation where that person never worked there, adjusted for a specific threshold of risk tolerance.
In other words, you can consider two hypothetical futures for a company: one with a specific person and one without that person. You then have a probability distribution defined over the difference in outcomes. For someone at a very high level, the absolute area under this curve is large—they have a big impact on the company (whether positive or negative). For someone at a lower level, their impact is small.
Someone who is good at their job should hopefully lead to positive impact, but you can certainly adjust this for risk. Perhaps you bring in a CEO who has a 90% chance of tripling the company’s revenue/growth and a 10% chance of leading the company to failure. That might be acceptable for a hypergrowth startup. For a larger and more mature company, you might have a different risk profile.
> A bit of a cynical take (on Hacker News no less) but after being in the industry for a while, my view is that the best definition of “level” is self-referential: it corresponds to the ability of a person to convince others that they are at that level.
I agree on paper that this is a great heuristic, but an employer would be more than happy to pay you a junior or mid compensation while extract senior level contributions from you. I see your point, though, and I agree for the most part.
> In other words, you can consider two hypothetical futures for a company: one with a specific person and one without that person.
This another tricky one - I agree on principal but I have seen this not work out in practice. For example, I have seen very senior people leave suddenly (like no notice period) and the business does not even skip a beat. In some cases, things get more efficient. My point being - employee level and comp sometimes do not map to contribution + business value.
That being said, I observed the above at big companies, where inefficiencies can hide in multiple levels of hierarchy and bureaucratic red tape. Its also why I hate working at big companies.
> an employer would be more than happy to pay you a junior or mid compensation while extract senior level contributions from you.
The problem with this argument is that the level also gets you a seat at tables you otherwise don't get invited to. The places where the decisions get made.
So "senior contribution" is not always possible without having the senior level on paper.
I have seen "senior" people hired from outside that do not make the same level of contributions as some one who is less senior by title and has years of tenure at their current position.
This is super common bc many companies would rather hire from outside instead of promoting from within. Also why there is so much turn over (engineers last for like 2-4 years then move on).
> I agree on paper that this is a great heuristic, but an employer would be more than happy to pay you a junior or mid compensation while extract senior level contributions from you. I see your point, though, and I agree for the most part.
In some ways, this actually makes sense to me. In my opinion, senior versus junior engineer seems like it should be about the skills and abilities of an individual, whereas one's compensation and actual title depend on a lot of environmental circumstances (company politics, for lack of a better term). While "soft skills" are an important part of a senior engineer, I'd argue that the set of soft skills useful to an individual contributor engineer doesn't necessarily overlap a ton with the soft skills needed to be able to effectively secure a promotion and/or higher compensation; navigating the corporate bureaucracy to the benefit of the team is primarily the responsibility of the manager. Finding a way to express the "level" of an engineer separately from their situation at their current company doesn't seem that crazy to me.
> Maybe that senior was great at training their replacement?
No, this was not the case. It turned out that those folks that left just did not contribute beyond superficial "management", delegating to subordinates, coordinating emails. Sort of like a "team secretary".
Ok, I'll guess I'll double-down since I am getting downvotes and it is still on topic.
I would be proud of someone said that about me. I have introduced new technologies (that work well) but I have also spent a significant amount of time training juniors, that I now know could handle the stack without me. Much because of their hard work of course, but a little bit because of me.
Maybe I was a one-trick-pony and this is all I knew. Or maybe I would continue improving the products by introducing new ways of working or implementing great stuff in the future.
Neither of this will be noticeable in a future without me.
But the worst kind of senior ought to be people that leave suddenly, leaving a chaos behind them.
It is like the "hero" who creates a chaotic product and puts a 100 hour week of bugfixing just before the release.
Compared to the person who just plans a product and executes according to schedule without leaving a mark.
> ...it corresponds to the ability of a person to convince others that they are at that level.
One note on this: as your career progresses, your ability to convince others around you that you're at a certain level goes from being unimportant to important, and then from there to essential.
After all, you need to influence others to do just about anything that involves more than yourself - and developing that power of influence is very much a skill in and of itself (see [1] for instance: it takes a lot to deliver even a simple, clear decision effectively!)
> Levels really shouldn’t factor into a problem discussion beyond determining who is involved in the discussion to begin with.
Titles shouldn't matter but they very often do. Question: "How should we integrate our product X with product Y?" Answer: "VIP David mandated that all integrations use technology Z, so it's already decided."
>I also have a definition of what I think level should ideally represent: the marginal contribution of a person’s influence on the outcome of a company, relative to the counterfactual situation where that person never worked there, adjusted for a specific threshold of risk tolerance.
Doesn't this tie back to the self-referential nature. If you can convince others of your "level", then you can also affect more change, making the level self-fulfilling, too.
Yeah, I hesitated to mention that as the post was already getting a bit long. I’ve always wondered what would happen if you took a random high performing junior engineer and just immediately gave them control over a large part of the company. Or vice versa—a level 8 engineer pretends to be a junior engineer and joins another company at level 3.
I suspect the degree of comfort and familiarity that people have in these situations has a lot to do with their potential to influence company outcomes. The junior engineer would still have very little impact when immediately given a VP role and likely proxy their decision-making through others at that level, but they would learn to adapt much quicker than someone going slowly from level 3 to 4 to 5, etc. (assuming the pressure didn’t cause them to quit). There’s a reason people say joining a startup accelerates career growth at the next company.
On the other hand, the level 8 engineer would immediately get themselves invited to discussions at a higher level, because they already have the social skills required to interact in those meetings. It’s not that most junior engineers can’t set up these meetings on others’ calendars—it’s that they don’t want to because they know they’ll stand out like a sore thumb in the discussions.
There’s definitely an element of luck involved with being in the right situations so you have the opportunity to make a larger impact, and once you do, you’re more comfortable with putting yourself in situations where you can do that again. So it is self-fulfilling in a sense. But there’s also a personality component, as more ambitious people (at least within the context of a corporate environment) are willing to take on more risk of embarrassing themselves by potentially failing at a higher level.
A "senior" cosplaying as junior can easily turn into conflict and deadlock when people feel their toes being stepped on by someone beneath them.
A junior cosplaying as senior reflects a situation with nepotism or favouritism. They are typically routed-around, quietly, by the people with relevant competence and jobs to do.
Real believable seniority is actually important in eliminating friction.
I get that, but even if titles were an amazing proxy for skill, that will only correlate with some subset of skills and say nothing about the rest (e.g. some staff engineers are amazing technically and only pretty good as team leads or vice versa). Even assuming that title correlates well with some notion of skill, there are organisations which won’t allow someone with Fancy Title to make a decision requiring skill C if they’re bad at it - and some would have no way of stopping them.
They've actually described the marginal revenue product of labor. In general, all of those counterfactuals fall under the concept of opportunity costs.
It probably doesn't matter as long as you keep it consistent. If the counterfactual was N-1 instead of a replacement worker, then almost everyone would represent positive value, and you're still comparing multiple positive numbers. Yes, there exist people who are net negatives and the company would be better off without them, but not THAT many.
I think there are many employees who are net negatives (who create less value than their full cost to the company). Sometimes this relates to the employee; many times it relates to the circumstance the employee finds themself in. (In almost any spiraling downward company, there are many employees in this situation, through no particular fault of their own.)
There are far fewer who are gross negatives (who contribute negatively overall, before compensation and other costs).
>I also have a definition of what I think level should ideally represent: the marginal contribution of a person’s influence on the outcome of a company, relative to the counterfactual situation where that person never worked there
What you are describing here is actual a stat they try to quantify in sports. W.A.R. Wins above replacement.
> In other words, you can consider two hypothetical futures for a company: one with a specific person and one without that person. You then have a probability distribution defined over the difference in outcomes. For someone at a very high level, the absolute area under this curve is large—they have a big impact on the company (whether positive or negative). For someone at a lower level, their impact is small.
Sounds a bit like the concept of "wins above replacement" as a baseball stat (https://www.mlb.com/glossary/advanced-stats/wins-above-repla...), which attempts to calculate how many fewer (or more, in the case of negative value players) a team would have if instead of a given player, they had someone producing the exact average of the league.
>A bit of a cynical take (on Hacker News no less) but after being in the industry for a while, my view is that the best definition of “level” is self-referential: it corresponds to the ability of a person to convince others that they are at that level.
This is right here is one of my biggest concerns at the moment. I know my technical skills are just as good as (if not better than) most of my peers with more YOE on paper, yet my communication skills are below them.
I really enjoy doing the technical stuff, but now I'm not so motivated to keep picking up technical skills because I've hit diminishing returns on it and instead now have to focus on something I don't like, nor have been great at for the longest time, and those are communication skills.
I do wonder if these skills are mutually exclusive to a large degree. ie, the thing that makes my peers better at communicating is exactly the same thing that makes less technical and vice versa. I worry about never being able to level up my comms skills without also taking my technical skills down a notch.
For me, I'm resigned to just looking for a new job to get a salary increase vs doing something I enjoy less, improving comm skills.
I think it is a trap to assume that communication skills and technical skills are mutually exclusive. In fact, many of the best engineers I've ever know were _also_ some of the best at communicating their ideas.
Granted, dysfunctional orgs, particularly large ones, will always end up promoting the sort of person I think you have in mind -- those who talk a great game, play politics, but don't really ship. That is an organizational and culture problem, though, and doesn't mean you should ignore your "soft" skills.
I think what you are looking for is what interests people.
Skills are not mutually exclusive but I suppose you are much more interested in technical stuff and to learn communication you have to become interested in other people.
I am also mostly interested in learning technical details of some software or system and not that much in what someone did with his time last weekend.
So what makes people better communicators is mostly that they are interested about other people and other people opinions and other people mood.
Yes, this is exactly what I'm getting at. In theory, those two are not mutually exclusive, but in practice they very much are due to personal preferences and that will shape what you become good at.
Of course, there are definitely people who excel at both, but they are outliers. Most will only get/want to excel in one of those two things.
This encapsulates a hunch I've had lately: that you could probably tell a lot about a potential hire if you could read their emails and slack messages.
The best co-workers are good at communicating without being annoying or a jerk, they have a good sense of what the important issues are, and they don't let ego or trivial concerns get in the way of solving problems.
Fwiw I was the CEO of my startup (obviously?). I could not even get why people are fighting each other, I was so naive and bad at communication.
Now I’m called a supercommunicator at the same company just because I took it seriously and took attention and vigorously checked out results of my communication attempts.
Yeah it was like 10 years, but first 1-2 with this mindset started with almost panicking to despair to be interesting to have more and more success. Now I almost enjoy it when I don’t think about how this is about fighting enthropy
and finding out who is silly in what way later to counteract that.
Might be even fun.
Edit: most fun part when you just became quite good and got it and the troublemakers are forgot you on the loser shelf, trying baby level politics stuff on you just because they are lazy. Yeah, then you can root them out and have a nice company.
I get were you are coming from. Focusing on comm skills is a big branch of a decision tree in terms future paths because it opens up possibility of joining the managerial class (CEO/CTO being potential options).
What convinced me against going down this branch for the time being is a quote from Naval Ravikant.
> No one can compete with you on being you. Most of life is a search for who and what needs you the most.
If I really enjoy tech stuff and not comms, I should focus on the tech stuff. Of course, I still need to have a good baseline at comms, but it doesn't have to be great.
I’ve been a senior engineer (definition loose) at a company I work at, and owner of a side business start up.
Politics at the startup are a bit easier, less than 10 people and such. But at the large company I’m trying to make sure I’m not caught off guard and be taken advantage of/other toxic politic stuff.
> A bit of a cynical take (on Hacker News no less) but after being in the industry for a while, my view is that the best definition of “level” is self-referential: it corresponds to the ability of a person to convince others that they are at that level.
This is of course not just in "the industry" -- for an extreme example there are a lot of elections arund the world in 2024 and almost all of them have at least one candidate trying to convince the hiring team (i.e. the voters) that they are qualified to do a job they've never done before.
> Perhaps you bring in a CEO who has a 90% chance of tripling the company’s revenue/growth and a 10% chance of leading the company to failure.
I realize these are hypothetical numbers, but they're so clearly backwards. It's more like a 10% chance for tripling the growth with a 90% chance of failure, and even that is being optimistic.
That depends on the changes proposed. Many places I have worked at could apply simple bits of internal automation and training that easily allows for trimming, or repurposing, half or more headcount. Keep in mind software is a cost center, so changes to tech teams will never result in greater revenue but they can greatly reduce expenses.
I also have a definition of what I think level should ideally represent: the marginal contribution of a person’s influence on the outcome of a company, relative to the counterfactual situation where that person never worked there, adjusted for a specific threshold of risk tolerance.
In other words, you can consider two hypothetical futures for a company: one with a specific person and one without that person. You then have a probability distribution defined over the difference in outcomes. For someone at a very high level, the absolute area under this curve is large—they have a big impact on the company (whether positive or negative). For someone at a lower level, their impact is small.
Someone who is good at their job should hopefully lead to positive impact, but you can certainly adjust this for risk. Perhaps you bring in a CEO who has a 90% chance of tripling the company’s revenue/growth and a 10% chance of leading the company to failure. That might be acceptable for a hypergrowth startup. For a larger and more mature company, you might have a different risk profile.