Hacker News Reader

Comment by tillmannhgl | Hacker News Reader

I think the idea of having three (or four or five) buckets of seniority is not applicable for modern work any more. Leveling is too much about hierarchy and too less about the actual impact. Beyond leveling is something what doesn't need to be defined in a word, but is more about specific knowledge or competence that is needed to solve a problem.

View on HN

It is an easy way to justify multiple pay bands.

I formerly worked for a company that for a few years referred to all of the engineers as "rock stars" etc. As the company matured a bit, that was dropped from all of the top-down type communication.

The company didn't actually value the employees any less, and we were still praised appropriately for good work and / or solid efforts. The issue was that people simply had to acknowledge that no, not all of the engineers put in the same level of effort or had the same natural talents, and pay disparities couldn't make sense when "everyone is a rock star programmer".

Granted, by the time i had left, we only had 3 or 4 pay bands from most junior to highest job title that still programmed, but the company wasn't in the west coast and didn't pay west coast salary either, so the actual compensation gradient was narrower than what you might find at a FAANG company.

Impact based leveling is deeply flawed: it privileges those who are assigned to work on the best products with the most legible success metrics. That’s how you get the kernel hackers who keep the lights on perpetually at entry level, and the people who were told to scaffold CRUD apps in the right place at the right time treated like rockstars. (I am the latter by the way).

This is why I like small organizations where compensation can be tied to impact.

In large orgs, there are so many positions that are important but have unclear or many levels removed from impact (HR, audit, sanitation).

I've also encountered many people who aren't motivated by the mission, but rather have other motivating factors outside of work and still want to progress. I think having some systematic way to include and grow people who don't have natural motivation for impact is important, for large orgs.

Sure it is, you gate underperforming people at the low bands and attrit them out.

That’s why in the regular army you don’t see 50 year old Captains.

So promote until the Peter principle is activated?

For the military... not really. They have a concept called “high year of tenure” which limits how long a soldier can remain in a rank without being promoted. You are meant to advance and if you don’t advance you will be separated out.

For example if you join the Army as a specialist (E4) you have 8 years to get promoted or you’re essentially kicked out.

Isn’t that the same thing? You eventually reach a level where you can’t perform at the next and become classed as incompetent.

Though I would think for enlisted this happens much less than officers.

No, you get promoted to level of competence, then fired after several years instead of becoming incompetent.

It would be the same thing if promotions weren’t so ludicrous. If you are an enlisted and want to get a promotion your best bet is to max your PT scores. It really has nothing to do with your ability to do a job at all.

Usually there’s a title standard at a particular band. A principal engineer is not just performing tasks for example, they are providing engineering/technical leadership.

Junior/Normal/senior is a change in rank or function. It makes sense to provide salary steps with respect to tenure, but not changes in duties.

I do wonder about the wisdom of this up or out (to some level usually) some times. For example, you can’t be SDE I at Amazon for 10 years. But why not? If I’m doing a good job as a SDE I, I’m happy with the job and pay and just want to keep doing it what’s the problem?

Because they want people who will grow that they can promote because it's really expensive to hire SDE III+

I don't think internally promoted SED III+ are any cheaper than externally hired ones. If they were they'd soon go else where anyways.

It's really expensive to hire at that level of seniority - even if the salary is the same once they start, it also takes a lot of time and money to get them to that point.

It’s hard to manage mediocre performers. Low performers you get rid of, and probably no one is too upset because even if they were well liked everyone recognizes that you can’t keep around low performers. High performers are obviously not an issue. But for mediocre performers, you risk the boiling frog where they slowly edge towards low performance but it’s hard to point to anything specific. And by sticking around forever they make a lot of allies and become increasingly hard to get rid of.

What about high performers that don't want to move up? I've known a few really good engineers that just don't want a lot of stress and don't care that much about the extra pay. They just want to do their work well at the level they are at and get paid.

I'm considered "senior" but the work in the last half of my career has certainly been more intense than the first half. Sometimes I think it'd be nice to go "back" a few steps to a job that I'm certainly very comfortable doing without too much effort and get paid decently well.

That’s definitely a downside to an up or out system. In other industries there’s a “mommy track” that allows for stepping back without stepping out but I don’t know of anything like that in tech.

Some companies see that as a risk.

I worked at a place with long tenured people who stick around forever. Worked great and saved money, but when they retired the business was pretty screwed.

You always have to bet on management being long term dumb, so processes need to be setup to maximize short term risk for management. No modern manager gives a shit about a high impact risk 5 years from now; if you’re lucky they may care about next year.

Reply on news.ycombinator.com