[HN Gopher] A Framework for Engineering Managers
       ___________________________________________________________________
        
       A Framework for Engineering Managers
        
       Author : aviramha
       Score  : 272 points
       Date   : 2022-07-28 07:15 UTC (9 hours ago)
        
 (HTM) web link (github.com)
 (TXT) w3m dump (github.com)
        
       | rishav_sharan wrote:
       | This is great. Is there something similar for Product Managers as
       | well?
        
         | ivalm wrote:
         | Isn't this TPM in this framework?
        
       | rgavuliak wrote:
       | The black labels on charts don't go well with the black GH
       | background
        
         | Sodman wrote:
         | Thanks for this, I didn't even see the labels and was re-
         | reading the text trying to see if they had specified which
         | chart directions represented which measurements!
         | 
         | Unfortunately GH doesn't have a way to alternate images in a
         | readme based on light mode vs dark mode afaik.
        
       | blowski wrote:
       | Interesting read. I'm not sure I agree that an EM defining
       | processes is operating at a more mature level than one merely
       | challenging them. I try to challenge processes but coach the team
       | to find the right processes themselves. I define processes myself
       | only as a last resort.
        
         | andsoitis wrote:
         | > one merely challenging them
         | 
         | I'm always more interested to hear from someone who _also_ has
         | suggestions for improvement or change.
        
         | aviramha wrote:
         | The way I see it - _you_ defined that the way to create further
         | definitions is by collaboration, so it 's still a way of
         | defining process. I assume it's not just a "democratic" process
         | to create a process ;) btw, totally agree on the point though!
        
       | buro9 wrote:
       | This attempts to recreate https://sfia-online.org/en , but it is
       | way harder than one imagines and then is in danger of being too
       | prescriptive and inapplicable.
       | 
       | It's better to go the other way and work out the principals
       | behind progression... that the further in their career the more
       | ownership, agency, and scope they have. Most things can be guided
       | by merely understanding the principals of it.
       | 
       | This very old doc (from Sun Microsystems engineers IIRC) remains
       | excellent
       | https://web.archive.org/web/20090420152505/http://mark.kampe...
       | 
       | It speaks to the spirit, the principals, behind progression...
       | and that makes it extremely applicable to lots of fields,
       | personal paths, and environments.
        
         | vladsanchez wrote:
         | Your point of "Ownership, agency, and scope" resonated with me
         | but it hasn't been my case. I have almost 30yrs experience and
         | still struggle to earn agency and ownership on subject-matter
         | topics/projects. Not sure what's my problem but my point is
         | that none of these frameworks are prescriptive, only
         | descriptive. Thanks for sharing your ideas.
        
         | chrisweekly wrote:
         | principals -> principles
         | 
         | (very different words)
        
           | buro9 wrote:
           | Yeah. Mobile phone keyboard are the bane of communication
        
       | fjdbeb76 wrote:
        
       | xEnOnn wrote:
       | It's a pretty comprehensive framework and I think it's great. It
       | would have been better if it also has a table of salary range for
       | each level in different regions so that companies aren't
       | misinterpreting these job levels or, at worse, misuse the guide
       | to manipulate their employees.
       | 
       | There are companies which make their employees senior or tech
       | lead but pay $50k and use the same framework as expectation. It
       | cannot be based entirely only on job titles because they are
       | "free" to issue. You can easily mint new titles out of thin air.
       | To back these job levels or titles with real value, their
       | corresponding pays need to be relatable.
       | 
       | If we could print money without backing it with something scarce
       | like gold, guess what will happen? Haven't we seen some companies
       | where there are so many senior, lead or even principal engineers
       | but yet they getting paid lower than a mid engineer at another
       | company? Job level/title inflation.
       | 
       | On the other hand, at companies that don't inflate their job
       | levels, they could still be underpaying their "seniors" and
       | "leads" with junior or mid level pay while expecting them to
       | operate along with the guide.
        
         | roguas wrote:
         | > If we could print money without backing it with something
         | scarce like gold, guess what will happen?
         | 
         | We do. We adjust monetary supply for markets to be stable -
         | money has been not backed by anything tangible for quite a
         | while. Like with any proposed framework(also applies to
         | monetary stuff) you find out how it affects outcomes and you
         | adjust usage.
        
           | xEnOnn wrote:
           | Yes, I'm aware that we do. I'm just trying to give an analogy
           | of what I'm trying to convey.
           | 
           | In the case of job levels, employees really shouldn't be
           | shortchanged into believing the job titles given to them. The
           | remuneration is a better indicator when in doubt.
        
       | nutate wrote:
       | wouldn't folks just join the military if this was their dream?
        
       | mosselman wrote:
       | This is great, thanks.
        
       | lamename wrote:
       | Reminds me of Patreon's IC leveling doc
       | https://levels.patreon.com/?
       | 
       | Note: '?' required for link to work
        
         | newfonewhodis wrote:
         | > Note: '?' required for link to work
         | 
         | Wow that is an extremely dumb feature (bug)?
        
       | lifeisstillgood wrote:
       | I think there is a cut off point where influence is entirely down
       | to hierarchical position.
       | 
       | Yeah there are some people who cross cut and have personal
       | influence (every org has celebrities) but in the main a person
       | who has 1000 people reporting to him has more influence than
       | someone with 2 people.
       | 
       | How they got to that position will have very little to do with
       | the skills sets mentioned here.
       | 
       | It's kid of like the French aristocracy before the guillotine was
       | invented writing a book on how to rise through the ranks to
       | become an influence at court.
       | 
       | It's missing most of the important stuff
       | 
       | PS - I am getting overly cynical but I do favour democracy in
       | organisations
        
       | zerkten wrote:
       | I wonder if the author plays Football Manager[1]. The attributes
       | screens use this same visualization.
       | 
       | [1] https://www.passion4fm.com/img/football-manager-player-
       | attri...
        
       | mariogintili wrote:
       | I feel like these structures are often used to prevent people
       | from getting a raise by raising the bar too far
        
       | Hermitian909 wrote:
       | This is really well done, but I feel like it (and other such
       | frameworks I've seen) all start to fall apart at high levels of
       | seniority.
       | 
       | What makes someone D7 in this framework? The real answer is not
       | in the charts, it's that they're very business critical. Maybe
       | they're some ultra-niche specialist or just the only one who's
       | been at the company since the beginning and knows where the
       | bodies are buried.
       | 
       | Things change in very interesting ways once the power dynamics
       | between employer/employee shifts so the company asking the
       | employee what it will take to make them stay.
        
         | aviramha wrote:
         | My take on management (and really on life) is that every
         | theory/framework is a setting stone and not instructions. The
         | framework helps you think better, but reality is always more
         | complex than what is described and covered in the framework.
         | Best managers IMO are aware of many different strategies,
         | frameworks etc and adapt it to suit _their_ organization needs
         | (based on the culture, people, goals, etc).
        
       | ChrisMarshallNY wrote:
       | Nice.
       | 
       | Like all of these types of things, I believe that an _heuristic_
       | approach is best, and this would work well for that.
       | 
       | I could see a training curriculum being developed from it.
       | 
       | However, it's also likely that an HR policy document would be
       | developed, with compensation and review targets. That could have
       | ... _interesting_ ... ramifications.
       | 
       | BTW: the image does not do so well in GitHub's new dark mode.
        
       | jb3689 wrote:
       | Interesting idea. Tech lead at our company is mostly a TPM role
       | mixed with the TL responsibilities listed here. In fact every
       | senior engineer is expected to act in the TL capacity at our
       | company. No commentary as to whether that is good or bad just
       | relating that is how our (fairly large, well-known) org works
        
       | webspaceadam wrote:
       | On the first glance i really like that approach, but i have the
       | feeling that something is missing. They address the issue, that
       | different people can be on different stages in different axes,
       | but there is no real explanation of how to handle the
       | differences. Someone can come in with little knowledge about the
       | technology, hence not using it before but is very well suited to
       | "lead" discussions about architecture or processes.
        
       | pliuchkin wrote:
       | Thanks, very useful. I'm just finding it hard do adapt and find a
       | place for the role of a Software/System Architect (which is very
       | common in my industry) in the framework since it conflicts a bit
       | with the other roles, mainly, Tech Lead. I`m thinking of
       | Software/System Architect to be a TL7.
        
       | mouzogu wrote:
       | thanks for this, very useful. at least a basic framework to work
       | with and adapt to my needs.
        
       | mepiethree wrote:
       | I really like this, the Dropbox career framework, and the Monzo
       | career framework. I synthesized all three recently into one that
       | I keep for our company, and the work pays such dividends. We do
       | quarterly performance reviews and not only does having a
       | consistent standard make these fair, but it reduced my time to
       | write them from 1 hour per report to about 30 min.
       | 
       | https://dropbox.github.io/dbx-career-framework/
       | https://github.com/monzo/progression-framework
       | 
       | And my nascent adaptation.. https://kevala-
       | progression.herokuapp.com/
        
       | harryf wrote:
       | It's a nice idea to things like this _in theory_. Medium has
       | something similar they called Snowflake -
       | https://github.com/Medium/snowflake ... but note the
       | disclaimer...
       | 
       | > Heads up: Medium isn't using this tool anymore, but you're
       | welcome to!
       | 
       | The problem with things like this is the typical engineering
       | mindset, when shown a graphical or numerical analysis of their
       | performance will zoom into analyse every detail and in the end
       | it's just backed by the managers impressions and judgement.
       | 
       | For similar reasons I have a strong opinion that engineers
       | shouldn't be paid performance related bonuses, because no matter
       | what happens, it's just going to upset them - the best case -
       | maximum bonus - just equates to a "meh". Better to pay a good
       | base salary upfront ...
       | 
       | Anyway use tools like this with caution IMO.
        
         | maest wrote:
         | > the best case - maximum bonus - just equates to a "meh"
         | 
         | Can you elaborate on this? I find it surprising since bonus
         | pays are very commonly used in many industries (and, in fact,
         | relied on very heavily in e.g. finance)
        
           | blowski wrote:
           | I remember going through performnace-related bonuses. If
           | somebody tried super-hard and got a "level 5", they got a
           | PS5K bonus. Somebody ticking along without being outstanding
           | got PS3K. Most people realised the extra PS2K wasn't worth
           | the extra effort required to get it.
        
             | pjbeam wrote:
             | The bonuses where I am now are a % of the employee's base
             | salary. A mid-level "meets" engineer can expect something
             | like $20-25k during an average year for the company. Up to
             | twice that for a strong year.
        
             | yakak wrote:
             | A raise of even a small amount makes a big difference over
             | time because raises continue and even compound a little.
             | Consequently, a cheap company will try to make the near
             | certain portion of a bonus feel like part of your
             | negotiated salary and extract effort with bonuses as
             | ephemeral raises.
        
               | quickthrower2 wrote:
               | It doesn't compound at all. If the going rate is $200k
               | and I switch jobs my previous salary doesn't matter, and
               | therefore compounding doesn't matter.
        
               | dzolvd wrote:
               | in what industry does your previous salary not matter?
               | (Not a snarky ask btw.) I have never entertained an offer
               | blow my current salary, so personally find them highly
               | correlated (sw eng).
        
               | pc86 wrote:
               | Nobody is talking about previous salary, the $3k/$5k
               | thing was about yearly performance bonuses.
        
               | jasongi wrote:
               | Previous salary can't affect future salary unless you let
               | the previous salary be known.
        
               | vikingerik wrote:
               | Raises that are a percentage will compound over multiple
               | iterations, that's what he meant.
        
               | pc86 wrote:
               | We're not talking about raises, we're talking about one-
               | time performance bonuses.
        
               | yakak wrote:
               | We are talking about performance bonuses, a trick to
               | avoid permanent raises.
        
           | romanovcode wrote:
           | It's different in finance because a finance guy can go to his
           | boss with hard proof and say "I made your company 10 million
           | last year" and ask for 500k bonus.
           | 
           | But developers do not have hard proof and they will be
           | getting 10k bonus instead.
        
             | HPsquared wrote:
             | See also, sales.
        
           | kqr wrote:
           | Performance bonuses the way they are commonly done doesn't
           | work in finance and sales either -- they're just so in-
           | grained in the culture that it's impossible to hire a finance
           | or sales person without offering bonuses the common way.
           | 
           | The problem is that individual performance is a much smaller
           | factor of outcome than commonly believed. Organisational
           | structure, luck, social standing in the organisation, etc
           | play a much bigger role. So in the end, performance bonuses
           | usually reward random variation more than individual prowess
           | (as much as the recipients would like to think otherwise.)
           | 
           | This one-armed bandit type compensation, when disguised as a
           | performance bonus, creates confusion, weird incentives,
           | limits creativity and experimentation, and breeds bad blood.
        
         | sokoloff wrote:
         | > Better to pay a good base salary upfront ...
         | 
         | Bonuses that are variable with company performance are a common
         | way to manage the reality for many businesses of "some years we
         | have more money than others". It seems better to pay bonuses
         | that vary than to set high base salaries and lay people off in
         | cyclical lean years.
        
           | kqr wrote:
           | That can work, if the scheme is (a) fair, and (b)
           | transparent. I.e. something like "this is the formula we
           | apply to the quarterly report to get everyone's bonuses".
           | 
           | Anything else just ends up being a lottery that helps no-one.
        
             | twblalock wrote:
             | > Anything else just ends up being a lottery that helps no-
             | one.
             | 
             | I've seen several occasions when corporate bureaucracy has
             | prevented or delayed raises and promotions because of
             | policy or budget wrangling, but not bonuses or stock
             | grants. Bonuses are a tool for managers to reward and
             | retain their high performers when those other options are
             | not available during that fiscal year.
        
         | rufius wrote:
         | I think your point is a good one. Though I'd say, that's why
         | it's good to describe these as frameworks. They set the rough
         | bounds of expectations, but much like API frameworks, the
         | business logic is specific to your company and its culture.
         | 
         | Anyway - agreed that you cannot just blindly follow one of
         | these. Also, this level of complexity in leveling is something
         | that is best left to bigger organizations. It's almost limiting
         | to impose it on smaller orgs.
        
         | charlie0 wrote:
         | I agree, pay them more. Nowadays, the best "bonus" is switching
         | jobs.
        
         | dasil003 wrote:
         | > _The problem with things like this is the typical engineering
         | mindset, when shown a graphical or numerical analysis of their
         | performance will zoom into analyse every detail and in the end
         | it 's just backed by the managers impressions and judgement._
         | 
         | The point of leveling frameworks is not to magically make
         | everything objective and quantifiable--though I agree that is
         | what many smart young people want and expect (after a decade+
         | in our Victorian education system)--it's to channel and provide
         | a rubric that many different engineering leaders' (both EMs and
         | ICs) impressions can be compared and contrasted in a structured
         | and relatively fair way. The last thing you want to do is
         | remove expert judgement from the equation, but in a medium to
         | large org it shouldn't be one person.
        
       | twinklegarg7 wrote:
        
       ___________________________________________________________________
       (page generated 2022-07-28 17:00 UTC)