Comment by johndoh42
Comment by johndoh42 4 days ago
Meanwhile the industry standard definition since the 80s:
- Junior - someone who can work under guidance.
- Regular - someone who can work alone.
- Senior - someone who can guide others.
Comment by johndoh42 4 days ago
Meanwhile the industry standard definition since the 80s:
- Junior - someone who can work under guidance.
- Regular - someone who can work alone.
- Senior - someone who can guide others.
If juniors ignore guidance and advice, they stay in junior roles, handling simpler, less impactful tasks.
Everyone seeks career growth, but pushing for it too quickly often just leads to inflated titles without real substance.
It’s perfectly fine to remain a mid-level engineer for your entire career if it makes you happy; it’s solid, honest work that contributes meaningfully. Plenty of people in their 60s have held the same job for decades, and that’s okay; it can be a path to genuine satisfaction.
Hinging senior evaluations on junior promotions directly fuels the title inflation I’m decrying. Desperate to show “impact through development,” seniors (or managers) push for premature title bumps; turning fresh juniors into “mids” or “seniors” without the skills to match, just to hit metrics.
This is rampant in tech, where inflated titles compensate for everything from low pay to talent wars, eroding expertise and making hiring a nightmare.
We end up with a system that prioritises optics over substance, where growth takes a backseat to checkbox promotions. It’s frustrating and counterproductive.
Mentorship should inspire organic development, not force-fed ladders that collapse under their own weight!
Instead, let’s measure seniors holistically, decoupling from junior title escalations to allow people to excel at their level indefinitely. Alternatives include:
* Technical Proficiency and Individual Contributions: Use code reviews, technical assessments, or metrics like deployment frequency and bug resolution rates to gauge a senior’s direct impact, without needing to “graduate” juniors.
This focuses on their own output and problem-solving prowess.
* Knowledge Sharing and Enablement: Track things like workshops led, documentation created, or peer feedback on guidance quality via 360 reviews—emphasising team uplift without mandatory promotions. * Project Outcomes and Efficiency: Evaluate based on team velocity improvements, innovation (e.g., patents or architectural wins), or overall delivery success, rewarding systemic contributions over individual mentee milestones. These methods honour diverse career paths, letting juniors stay put if it suits them while still valuing (and evaluating) senior leadership.
Agreed, a direct metric of “promotion rate” is obviously flawed. I posed it more as a rhetorical question for reflection- at the limit, it’s clear that people with mentoring responsibilities should be accountable somehow for being good mentors. Terrible mentors who undermine, sabotage, or consistently fail their mentees definitely exist.
A lot of the comments are complaining about how this metric is a terrible way to evaluate seniors but I'd disagree. If one junior can't grow then it's a problem for that junior. If no juniors can grow then it's a problem for the senior - either they don't have good mentoring skills OR they need to work on improving the hiring pipeline for their team (either raising the bar in interviews/changing what skills you're evaluating for in interviews/working with recruiters to fix the pre-interview filters).
In either case it's an ambiguous problem that needs to be solved and just throwing your hands up and saying that you don't want to be evaluated for that is not going to help.
I've never worked at a place like that, and I hope I never do. Although, I've never even heard of any reasonable person putting that into practice either, so I'm probably safe.
> Any mentor type figure is going to be at least partially evaluated by progress of the mentees against some benchmark.
Sounds like the same kind of mistake as evaluating teachers by the grades of their students. Soon people figure out the "one weird trick" how to get the highest score easily.
People perform to your metrics. If you don’t want people to be one trick ponies, you’d better have more than one metric.
> Everyone seeks career growth, but pushing for it too quickly often just leads to inflated titles without real substance.
That's why I'm not a big fan of recommending people to often and quickly change jobs to increase titles and pay. Their skills don't level up the same way, and they end up with a title of senior/lead developer and can't actually build maintainable systems or solve problems that nobody tells them the solution to.
Agree.
If one is unable to work alone but manages to join a new company with an inflated title, people will notice. They're gonna have to keep job-hopping until they find a place that doesn't notice the bad performance anymore.
This is demonstrable by the amount of CVs with "12 jobs in the last 6 years" in my reject pile.
talk to their manager. If their manager doesn't respond you go to your manager or the manager's manager.
Yes but there is also a temporal component as well. A Senior should be able to do all their tasks and whatever else comes their way without needing guidance. To be able to do that requires a certain level of time in position.
I do wonder how seniors manage cultural / technical differences. If the junior is not responsive to guidance, advices, hints .. what else do you do