> One caveat: my experience comes mainly from working on infrastructure and developer tools at large companies, on teams where engineers have a lot of bottom-up autonomy to influence their roadmaps. In a more top-down environment, there may simply be less room to work this way.
I wonder if the overall trend in tech is that engineers are experiencing less bottom-up autonomy and more top-down controlled environments. I would be curious to see how many tech companies (or the average engineer's experience) have changed from being tech led where engineers have autonomy to being more product management led. My suspicion without evidence is that the overall engineering autonomy has decreased over the years as the culture of tech has (in my opinion) shifted away from tech focus to more business, management, product focus with engineers just as the widgets who are tasked with fulfilling the goals of business, management, product.
Without providing too much detail, at my current place of employment I find that to be the case.
In most of my career, especially when in a Staff role, it has been up to me to propose the direction and priority of work within my scope. Usually, this is based on my estimation of the impact to the business (possible new features, security) or operations (devx, efficiency $$$). Obviously I still have to work with product to get things scheduled as they have needs that must be met, but it was as an equal partner. As a very creative person thats usually the space I enjoy the most.
I agree with this and think that when your company isn't profitable and is running out of runway (due to the end of the zirp) and can't get more, moving toward features that more directly could land sales is a natural top-down move (and totally correct in the abstract).
However I still see a lot of "work theater" where directors couldn't care at all about changing the bottom line but just care that work that sounds plausible enough is executed in a predictable way that makes them look good (in fact they get pretty disappointed if something happens TOO fast [like half the estimate] as it illustrates they are out of touch the work they are claiming to represent). What I'm saying is that "do more top-down-product" in theory makes sense as a direction, but in practice I usually see it as wasted energy.
Most of my career has been spent in JavaScript and then MuleSoft. I have only ever seen top-down authority in about 20 years of doing this, except for when I was at Travelocity early in my career.
The general perception is that JavaScript work in the corporate world is for young people who have no idea what they are doing and lack the maturity to write original solutions. This perception is common from developers who do other work, management, and developers who primarily perform JavaScript work. If I want to be creative or have autonomy to do anything more ambitious than putting text to screen I have to save it for personal side projects or do unrelated work.
MuleSoft work was just more of the same with all decisions funneled to strict silos of a select few and everybody else was just a keyboard pressing stooge.
Yes, that does sound dreadful, but what's worse is that it amplifies the personalities of people who will throw others under the bus for attention and potential rock stars who perform their most diligent work at hiding from that attention.
I think the autonomy was stripped not that long ago and it's making software products far worse.
It doesnt make sense that this would happen from an economic perspective at first glance - not unless you consider labor (devs), execs and investors to be three competing groups who are vying for power.
I worked at the Staff level for a while, I think the key is to report to a director who manages managers. Every time I did I had a good time, projects were easy to define and people were easy to convince. I had the same level with a few managers who were managing other ICs, and that never worked. Other ICs try to compete on tasks, people don't come to you with their needs, you learn about different projects often too late, other teams are territorial and don't really want you to intervene.
The part I find challenging in the senior-staff twilight is that having deep technical knowledge means I can solve short term problems, just as requests, fast and effectively.
The author mentions that you should spend time understanding the frustrations from other teams, but that takes up loads of time and I don’t like being the person who talks and talks but doesn’t push code and ship features. I’d love to hear how others have experienced that
I’ve experienced this. The critical realization for me was that most of those problems which I know I can solve quickly, without even having to write a Jira ticket or groom them into a sprint or whatever, are problems I should let someone else solve.
Sure, I can do them faster than others and pretty well. But if they’re truly things I can knock out in a day or two, they’re things that other staff can knock out in a week or two, and learn from the experience, and at that size they likely don’t require the standard of quality that I delude myself that I hold myself to.
At the lead/staff level, it’s far better to look for the kind of problems described in TFA: problems that are complex to identify, and for which the appropriate solutions aren’t always the obvious ones. My time is spent much better looking for those than being a 10x-speed senior engineer or whatever. There are actual senior engineers for that; I’m paid for the kind of work that they don’t or can’t do (yet, and to help model and mentor and train them to be able to do it), not to do the kind of work they already can do faster/better.
It’s case-by-case of course; sometimes a simple issue is critical or obscure enough (or tempting enough to override my iffy-at-best self-discipline) that I’ll jump on it. But I generally try to remember that a lot of the stuff I could look heroic for fixing in a jiffy is probably both not worth my salary allocation in the eyes of my grandboss, not critical time-wise, and a potential learning/accomplishment opportunity for others.
Funny people should have that problem. I have spent most of my career in the startup space, and my experience has consistently been that the amount of problems to solve is vastly larger than what I can reasonably achieve in my waking hours.
So I don't find problems to solve, I try to assess which problems are most urgent, or which solution solves several of them at once. Learning to get that kind of prioritisation right to keep all teams and customers happy and productive is what I'm very proud of in my career.
There are a lot of known problems... but there are also a lot of unknown, or at least unrecognized problems. Some of those unrecognized problems are going to be more useful and higher-leverage than recognized problems. The same "being a sponge" approach still helps.
I am doing exactly same thing of creating a solution which can be applied to multiple problems at the same time.
Sometimes coworkers does not like that because they don't see me working on the problem, but on a tool which will resolve the problem and similar problems from our backlog.
It’s the same in a large corporate environment. I have a “personal projects” doc of ideas I’ve had to make development experiences better at my current role - it’s a couple hundred lines long.
I managed to get a couple of smaller tools out recently thanks to having copilot available to churn on them while I spend my time on prescribed work, but it would consume all of my time to even make a significant dent in it.
He says: That taught me to let potential problems pile up. Listening the way I do leaves me with far more of them than I could possibly solve, and not all deserve action. Most don’t need to turn into projects the first time I hear about them; waiting can be a superpower.
Yes. He also says: “How do you find problems worth working on?” a senior engineer I mentor asked me recently. My point is that I'm working in a whole different environment, and the challenge - to me - is never finding interesting problems to solve, but identifying the most important problem out of a large pool of known problems.
I think almost all of tech is bloated and massive layoffs wouldn’t phase most companies (though it would be harmful to people’s lives so it seems cruel to do this). Fewer people per teams means less context switching and devs own more. They don’t have to look for work, it will be in front of their face. In so many of these big tech companies I’ve worked at I’ve seen too many not having enough work to do. They end up creating meetings and other wasteful things (doc writing) to occupy their time. Managers and directors seem to want big bloated teams, the larger their head count the more they can demand and push for a promo.
yea this was the case even before ai. fighting for scope is 80% of the job now. Ppl like OP dont really need to "Act like a sponge" and figure out "problems to work on" if there is natural demand for stuff they are building work will present itself.
There is also crazy ladder climbing with all these titles and difference in payscale. So ppl like the author are trying to optimize what a role X is supposed to be doing. Everyone is just the same thing everywhere.
Everywhere you go its the same ppl, same ladder , same tools same sucking up to boss blah blah. Cant wait to get out of this shit.
One thing that isn't often mentioned in these discussions:
Making sure that the work was actually done.
I've been a Staff Engineer and managed engineers and it's pretty shocking what people consider to be "done".
A couple examples:
- Doing a migration to using Tailscale and an engineer claims it's done even though there is just one giant ACL for the whole firm
- Migrating from one monitoring system to another despite only 80% of the alerts have been migrated
- etc
Some of this is business folks creating bad incentives. Some of it is not creating good "success criteria" for projects. Either way, someone has to go through and make sure that both the details and the big picture deliverables landed correctly.
This being HN, I'm sure someone will say something like "just hire better engineers". I've seen phenomenal engineers make bad choices here due to poor incentive design.
A perfect example:
- you reward people for hitting delivery deadlines
- you punish people when there are outages
you might think you're pretty smart until you realize the odds of getting yelled at if you miss a deadline is 100% but the odds of an outage are <100%. The EV+ outcome then becomes to hit the deadline even if you know the code isn't ready.
Again, the job of senior engineers/engineering managers is to make sure these things don't happen by both double checking work and also pushing back on bad incentives.
My advice as someone working at staff engineer role for 3 different employers remotely and the "let problems accumulate" resonates but different context:
- don't be too proactive in solving issues that signals you are not busy to your employers.
- you are not hired to sit around 9-5, what you ship and how it impacts the business bottom line is far more important.
Especially true when I have to juggle 3 different employers. I will be in a long standup meeting with company A while I am answering slack messages from company B or company C has a deadline that overlaps with another and I'd have to work on at the same time.
Previously without LLMs this was very difficult but now its manageable, especially with openclaw and hermes doing a lot of lifting. This let me discover a lot of what's discussed in the article naturally.
Thanks for letting me know; I cannot reproduce this myself on my mobile (it should be white text on a dark grey background) but I pushed a speculative fix which might improve things. Please let me know if it helps or, if not, I would love to know more details so I can fix this!
Listen to your customers but ignore what they say. “Treat your customers like a kindergarten class. If they’re all asking for a snack they’re hungry but perhaps they really need a nutritious lunch instead.”
I spent the early part of my career (mid 2000s) working on a registration system for sports tournaments.
I had two jobs:
1. Writing the actual software for the registration website
2. Going to the events and managing the registration tent where people actually checked in
Doing 2 gave me a VASTLY better understanding of who the users were and how they thought. e.g. I assumed it would be tech savvy, organized people like me in their mid 20s. It was actually a mix of 40 something team dads who weren't tech savvy at all (b/c mid 2000s) but very organized and teenagers who were the opposite.
It meant effectively managing two completely different customer bases in the same product.
Coupled with the fact that cable modems were a new thing but not evenly distributed meant that we had to design the site accordingly.
At the time, I remember Joe Spolsky mentioning that at Microsoft they did two way mirror usability testing and it totally made sense. I wish more firms did that today.
Eh, I'm not going to say this is wrong, but pretty much like all advice, it's kinda cheap without data. Like Dale Carnegie may be very famous and have a book and say "Use people's names all the times and your conversations will be great" but who actually knows if that makes a difference without some controlled outside study.
I see a lot of people really eager to tell other people how to staff, but a lot of them sure do seem to disagree, and most people I know get to staff anyways without any particular skill after a certain age (title inflation?) including myself.
My smell test for an engineer is -- are they trying to quantify the size of everything in terms of impact, or are they just reacting to whatever customer knocks on their door?
The author notes:
> One caveat: my experience comes mainly from working on infrastructure and developer tools at large companies, on teams where engineers have a lot of bottom-up autonomy to influence their roadmaps. In a more top-down environment, there may simply be less room to work this way.
I wonder if the overall trend in tech is that engineers are experiencing less bottom-up autonomy and more top-down controlled environments. I would be curious to see how many tech companies (or the average engineer's experience) have changed from being tech led where engineers have autonomy to being more product management led. My suspicion without evidence is that the overall engineering autonomy has decreased over the years as the culture of tech has (in my opinion) shifted away from tech focus to more business, management, product focus with engineers just as the widgets who are tasked with fulfilling the goals of business, management, product.
all hypothesis, only anecdata
Without providing too much detail, at my current place of employment I find that to be the case.
In most of my career, especially when in a Staff role, it has been up to me to propose the direction and priority of work within my scope. Usually, this is based on my estimation of the impact to the business (possible new features, security) or operations (devx, efficiency $$$). Obviously I still have to work with product to get things scheduled as they have needs that must be met, but it was as an equal partner. As a very creative person thats usually the space I enjoy the most.
I agree with this and think that when your company isn't profitable and is running out of runway (due to the end of the zirp) and can't get more, moving toward features that more directly could land sales is a natural top-down move (and totally correct in the abstract).
However I still see a lot of "work theater" where directors couldn't care at all about changing the bottom line but just care that work that sounds plausible enough is executed in a predictable way that makes them look good (in fact they get pretty disappointed if something happens TOO fast [like half the estimate] as it illustrates they are out of touch the work they are claiming to represent). What I'm saying is that "do more top-down-product" in theory makes sense as a direction, but in practice I usually see it as wasted energy.
Most of my career has been spent in JavaScript and then MuleSoft. I have only ever seen top-down authority in about 20 years of doing this, except for when I was at Travelocity early in my career.
The general perception is that JavaScript work in the corporate world is for young people who have no idea what they are doing and lack the maturity to write original solutions. This perception is common from developers who do other work, management, and developers who primarily perform JavaScript work. If I want to be creative or have autonomy to do anything more ambitious than putting text to screen I have to save it for personal side projects or do unrelated work.
MuleSoft work was just more of the same with all decisions funneled to strict silos of a select few and everybody else was just a keyboard pressing stooge.
Yes, that does sound dreadful, but what's worse is that it amplifies the personalities of people who will throw others under the bus for attention and potential rock stars who perform their most diligent work at hiding from that attention.
I think the autonomy was stripped not that long ago and it's making software products far worse.
It doesnt make sense that this would happen from an economic perspective at first glance - not unless you consider labor (devs), execs and investors to be three competing groups who are vying for power.
I worked at the Staff level for a while, I think the key is to report to a director who manages managers. Every time I did I had a good time, projects were easy to define and people were easy to convince. I had the same level with a few managers who were managing other ICs, and that never worked. Other ICs try to compete on tasks, people don't come to you with their needs, you learn about different projects often too late, other teams are territorial and don't really want you to intervene.
The part I find challenging in the senior-staff twilight is that having deep technical knowledge means I can solve short term problems, just as requests, fast and effectively. The author mentions that you should spend time understanding the frustrations from other teams, but that takes up loads of time and I don’t like being the person who talks and talks but doesn’t push code and ship features. I’d love to hear how others have experienced that
I’ve experienced this. The critical realization for me was that most of those problems which I know I can solve quickly, without even having to write a Jira ticket or groom them into a sprint or whatever, are problems I should let someone else solve.
Sure, I can do them faster than others and pretty well. But if they’re truly things I can knock out in a day or two, they’re things that other staff can knock out in a week or two, and learn from the experience, and at that size they likely don’t require the standard of quality that I delude myself that I hold myself to.
At the lead/staff level, it’s far better to look for the kind of problems described in TFA: problems that are complex to identify, and for which the appropriate solutions aren’t always the obvious ones. My time is spent much better looking for those than being a 10x-speed senior engineer or whatever. There are actual senior engineers for that; I’m paid for the kind of work that they don’t or can’t do (yet, and to help model and mentor and train them to be able to do it), not to do the kind of work they already can do faster/better.
It’s case-by-case of course; sometimes a simple issue is critical or obscure enough (or tempting enough to override my iffy-at-best self-discipline) that I’ll jump on it. But I generally try to remember that a lot of the stuff I could look heroic for fixing in a jiffy is probably both not worth my salary allocation in the eyes of my grandboss, not critical time-wise, and a potential learning/accomplishment opportunity for others.
Funny people should have that problem. I have spent most of my career in the startup space, and my experience has consistently been that the amount of problems to solve is vastly larger than what I can reasonably achieve in my waking hours.
So I don't find problems to solve, I try to assess which problems are most urgent, or which solution solves several of them at once. Learning to get that kind of prioritisation right to keep all teams and customers happy and productive is what I'm very proud of in my career.
There are a lot of known problems... but there are also a lot of unknown, or at least unrecognized problems. Some of those unrecognized problems are going to be more useful and higher-leverage than recognized problems. The same "being a sponge" approach still helps.
I am doing exactly same thing of creating a solution which can be applied to multiple problems at the same time.
Sometimes coworkers does not like that because they don't see me working on the problem, but on a tool which will resolve the problem and similar problems from our backlog.
This is not distinct.
Even in a big corp there are always problems to solve. The point is to find problems that both:
1) are hard enough that someone junior won't be able to solve them alone, and
2) actually provide value to the company / users
You probably pivoted your career towards the startup space so you found it effortless. For others it is basically a castle with a 20m wall.
It’s the same in a large corporate environment. I have a “personal projects” doc of ideas I’ve had to make development experiences better at my current role - it’s a couple hundred lines long.
I managed to get a couple of smaller tools out recently thanks to having copilot available to churn on them while I spend my time on prescribed work, but it would consume all of my time to even make a significant dent in it.
He says: That taught me to let potential problems pile up. Listening the way I do leaves me with far more of them than I could possibly solve, and not all deserve action. Most don’t need to turn into projects the first time I hear about them; waiting can be a superpower.
Yes. He also says: “How do you find problems worth working on?” a senior engineer I mentor asked me recently. My point is that I'm working in a whole different environment, and the challenge - to me - is never finding interesting problems to solve, but identifying the most important problem out of a large pool of known problems.
“Can be a superpower” is so empty. What does that mean outside of trying to proliferate some cliche?
Just means it can be powerful or useful technique that can feel “magical” when employed.
I wouldn’t overthink this.
Not overthinking can be a superpower, but writing eloquently can be a superpower, and not trailing cliches behind your pen can be a superpower.
I think almost all of tech is bloated and massive layoffs wouldn’t phase most companies (though it would be harmful to people’s lives so it seems cruel to do this). Fewer people per teams means less context switching and devs own more. They don’t have to look for work, it will be in front of their face. In so many of these big tech companies I’ve worked at I’ve seen too many not having enough work to do. They end up creating meetings and other wasteful things (doc writing) to occupy their time. Managers and directors seem to want big bloated teams, the larger their head count the more they can demand and push for a promo.
yea this was the case even before ai. fighting for scope is 80% of the job now. Ppl like OP dont really need to "Act like a sponge" and figure out "problems to work on" if there is natural demand for stuff they are building work will present itself.
There is also crazy ladder climbing with all these titles and difference in payscale. So ppl like the author are trying to optimize what a role X is supposed to be doing. Everyone is just the same thing everywhere.
Everywhere you go its the same ppl, same ladder , same tools same sucking up to boss blah blah. Cant wait to get out of this shit.
One thing that isn't often mentioned in these discussions:
Making sure that the work was actually done.
I've been a Staff Engineer and managed engineers and it's pretty shocking what people consider to be "done".
A couple examples:
- Doing a migration to using Tailscale and an engineer claims it's done even though there is just one giant ACL for the whole firm
- Migrating from one monitoring system to another despite only 80% of the alerts have been migrated
- etc
Some of this is business folks creating bad incentives. Some of it is not creating good "success criteria" for projects. Either way, someone has to go through and make sure that both the details and the big picture deliverables landed correctly.
This being HN, I'm sure someone will say something like "just hire better engineers". I've seen phenomenal engineers make bad choices here due to poor incentive design.
A perfect example:
- you reward people for hitting delivery deadlines
- you punish people when there are outages
you might think you're pretty smart until you realize the odds of getting yelled at if you miss a deadline is 100% but the odds of an outage are <100%. The EV+ outcome then becomes to hit the deadline even if you know the code isn't ready.
Again, the job of senior engineers/engineering managers is to make sure these things don't happen by both double checking work and also pushing back on bad incentives.
My advice as someone working at staff engineer role for 3 different employers remotely and the "let problems accumulate" resonates but different context:
- don't be too proactive in solving issues that signals you are not busy to your employers.
- you are not hired to sit around 9-5, what you ship and how it impacts the business bottom line is far more important.
Especially true when I have to juggle 3 different employers. I will be in a long standup meeting with company A while I am answering slack messages from company B or company C has a deadline that overlaps with another and I'd have to work on at the same time.
Previously without LLMs this was very difficult but now its manageable, especially with openclaw and hermes doing a lot of lifting. This let me discover a lot of what's discussed in the article naturally.
I think the first problem may be fixing the broken mobile experience. I'm getting black text on a very dark gray background.
Thanks for letting me know; I cannot reproduce this myself on my mobile (it should be white text on a dark grey background) but I pushed a speculative fix which might improve things. Please let me know if it helps or, if not, I would love to know more details so I can fix this!
Thanks!
Listen to your customers but ignore what they say. “Treat your customers like a kindergarten class. If they’re all asking for a snack they’re hungry but perhaps they really need a nutritious lunch instead.”
I spent the early part of my career (mid 2000s) working on a registration system for sports tournaments.
I had two jobs:
1. Writing the actual software for the registration website
2. Going to the events and managing the registration tent where people actually checked in
Doing 2 gave me a VASTLY better understanding of who the users were and how they thought. e.g. I assumed it would be tech savvy, organized people like me in their mid 20s. It was actually a mix of 40 something team dads who weren't tech savvy at all (b/c mid 2000s) but very organized and teenagers who were the opposite.
It meant effectively managing two completely different customer bases in the same product.
Coupled with the fact that cable modems were a new thing but not evenly distributed meant that we had to design the site accordingly.
At the time, I remember Joe Spolsky mentioning that at Microsoft they did two way mirror usability testing and it totally made sense. I wish more firms did that today.
[flagged]
[dead]
Eh, I'm not going to say this is wrong, but pretty much like all advice, it's kinda cheap without data. Like Dale Carnegie may be very famous and have a book and say "Use people's names all the times and your conversations will be great" but who actually knows if that makes a difference without some controlled outside study.
I see a lot of people really eager to tell other people how to staff, but a lot of them sure do seem to disagree, and most people I know get to staff anyways without any particular skill after a certain age (title inflation?) including myself.
My smell test for an engineer is -- are they trying to quantify the size of everything in terms of impact, or are they just reacting to whatever customer knocks on their door?