Writing a thread on onboarding here, because I don't want to write it in the other place.
Context:
Good engineer onboarding *is* DEI. Teams that have great onboarding, tend to be more diverse and inclusive. Teams that have poor onboarding, tend to be more homogeneous and less diverse.
That's because poor onboarding makes life harder for junior employees and remote employees. Underrepresented groups benefit disproportionately from remote work options. Junior employees are a more diverse group.
Poor onboarding leads to more "beginner questions" for longer. Underrepresented groups experience higher social threat. So if 3 employees start working at Tech Corp on the same day, and they are:
* A white man
* A white woman
* A Black man
And they all have the same beginner questions 6 months after their start date, then the Black man and white woman are more likely to be perceived as "slow ramp up" people, and face career consequences as a result.
Imposter syndrome isn't real. Imposter phenomenon is very real.
Imposter phenomenon is the combination of two things:
1) the belief that I know less things than others around me.
2) the belief that if this is found out, that I will experience harm from it.
In short, it's about psychological safety. It's a rational response based on the lived experience of the person feeling it, and their observation of how others are treated, it's not a shortcoming of the person experiencing it.
There are lots of things that we can do to minimize the imposter phenomenon. Better, more consistent, and deterministic ramp up experience, is one of these things. 👍🏿
This doesn't just mean documentation. This also means removing processes that shouldn't exist in the first place.
But in terms of documentation, one thing that really helps, is having a verifiable definition of being ramped up and onboarded.
Welcome Starfleet graduate! You've been assigned to the Transporter team! "Dream to Beam!"
Week 1:
✅ Get desk, comms badge, quarters
✅ Meet your onboarding guide (Scotty)
⬜ Join all team mailing lists
⬜ Take the "Teleportation basics" training
Month 1:
⬜ Complete first pair transport (inanimate objects)
⬜ Complete first solo transport (sentient life forms)
⬜ Join duty rotation (shadow your guide)
Quarter 1:
⬜ Complete first multi-party transport
⬜ Join duty rotation (solo)
⬜ You're onboarded!
The other thing about intentional, verifiable ramp-up plans, is that they make the person ramping up less nervous about their own progress.🤯
Without the plan:
"I've been here 20 days, and I haven't even transported a living thing yet! Not even a Tribble! I was top of my class at Starfleet, and I thought my paper on "Subspace tunneling to safely beam matter through singularities" impressed them! Do they not trust me? Am I messing up in other ways? Is... is it because I'm half Romulan?!"
One other thing we've noticed that makes a huge difference, having a deliberate "Onboarding buddy" someone that the new hire can ask questions. Even better if they're not a manager (less concern about looking dumb).
Not knowing who to ask can be paralyzing for some. Having someone senior they can ask and know they're not bothering, removes lots of barriers. Plus, it's good experience for the senior person.
Did @danhon write this onboarding checklist?
heh, kinda happy that the onboarding template looks an awful lot like this.
I cleaned it up for public use. Not sure if it actually is useful, but yeah, making things easier for anyone to onboard is A HUGE WIN.
https://drive.google.com/file/d/1hWEgOYrH07-t_w-hPOYjEp-ooj1mxogv/view?usp=sharing
(Don't make new folk suffer like you did. They don't need to earn anything. You hired them to help free up your work.)
In the flip side, I offered comments for ours and was met with "we've been doing it this way for 25 years, it's fine" 😂😒
Apologies for a billion notifications -- something about my keyboard went nuts.
Re the "get reading!"... is there a navigation path through the voluminous (right?) docs? A prioritized reading list or something?
@mekkaokereke Part of the reason I left academia years ago is my "onboarding" process for the first few months of my first post-Ph. D. job looked like this:
Week 1:
[X] Get desk, university ID card, keys to the lab, and stuff like that
[X] Start learning in more detail about projects the lab does
Month 1:
[X] Pick a project.
[X] Attend group meetings.
[X] Learn how to use some of the lab's instrumentation.
Quarter 1:
[] Be familiar with all lab instrumentation including the ones with no manuals
[] You’re still asking other people for help with this? Cut it out. You should know how to do all of this by now.
[] Be able to undo previous people's experiments before doing your own, without fscking them up
[] You’re still asking other people how to avoid screwing up their stuff when you set up your work? See above about knowing how to do all that by now.
[] Have enough data on your chosen project to start presenting meaningfully in group meetings
[X] Project not working? Sucks to be you. Find another one to get more data from.
Half-year 1:
[] You have lots and lots of good data, right?
[X] OK, observe your PI kick collaborating prof out of collaboration and saying we're going to go our own way
[X] ...Including managing all the supplies, lab protocols, etc., that the collaborating team was the expert on, not you
Year 1:
[] Start writing first paper.
[] Congratulations, you're onboarded!
[X] Oh, no useful data? Pull trigger on "leave not just job but career and ponder something less stressful, like software development at a small company that wants all things web, right now"
By comparison, onboarding as a software developer/engineer has been much less painful.
Replace "ask Bob for an a/c on the VCS repo" with "just have sso everywhere."
This!
The provocative question is:
"Engineers, could a principal engineer from say Microsoft, join your team, and without talking to anyone, just by reading docs, fix a trivial bug in your system?
And push it to production?
On their first day?
Not saying this should always be possible. Just baselining onboarding."
When we think of the reasons why this might be "no," it usually comes down to things like ACLs, permissions, non-documented knowledge, lack of CI/CD, etc.
Yes, absolutely. We aimed for, and achieved that, at $job-2.
It's really basic stuff that demonstrates you value someone's contribution, want them to feel able to get stuck into stuff, and have trust in yr own systems to enable that w/o toasting some production kit.
Anything less is a trifle lacking, if I'm feeling brutally honest.
@mekkaokereke @JuliaRez @francois one of the hardest parts of this is for anyone that's never done it it sounds impossible.
90% of the work is convincing them that something like this is an achievable realistic goal.
Like, if you've only ever worked with bad managers you have no frame of reference for what good management looks like.
@nikclayton @mekkaokereke @JuliaRez @francois This isn’t just bad managers, it’s also institutional bureaucracy and inertia. “We’ve always done it that way”. “That’ll get shot down by the Security Team”.
Having just joined a .edu, my leadership has been great and try to help, but they are also oversubscribed and only have so much time from existing commitments. Now you got some new guy fresh from industry talking about “All fit repos should be visible” and “Service Account tokens should be usable by all environments for an application”[*] or “we really should remove or refactor this part of the pipeline because the guy who wrote and maintained it left five years ago and the other guy who learned it is retiring next year.”
They agree with the changes, but the disruption/downtime risk is more than they have in their risk budget will allow.
[*]: Each new token requires a pull request from the auth team and that can take up to a day to get merged.
@nikclayton @mekkaokereke @francois
Crikey yes. Institutional inertia is _weird._ You can have the set of devs + ops types on-side, and mgmt nodding along at the thing they don't understand... And yet. It's like some sort of ichor of awful seeps from the walls and poisons the good intentions.
Sometimes, if you're a white cis dude with a deal of social capital within the org, you can steamroller that.
Otherwise, not.
(I am one of the people who got to A/B test that)
💯 That's a powerful tool that I've used in the past to combat the imposter phenomenon.
This addresses the "I don't know so many things that other people know!" If a junior SWE becomes the person at the company that knows the most about [feature X], and they see *very* senior SWEs from other teams, come to them and ask them questions, their thinking becomes:
"Yes, they know things that I don't know. Facts. But I know things that they don't know! No one knows it all! This is normal!"
@mekkaokereke I wonder also about the concept of “beginner questions” I’m sure many times they are basic stuff that should have been covered in onboarding but aren’t (because the people may assume they are obvious because they have forgotten that not everyplace works the same way or uses the same language)
But equally I could easily imagine some are cases of people with an outside of the org perspective seeing alternatives perhaps from their own past experience and asking questions first
@mekkaokereke I came to a realization that my imposter syndrome (or phenomenon) was rooted in my undiagnosed ADHD.
Once I admitted to myself that I had spent my whole life fighting a developmental disability on my own (and being punished for failing), it helped a lot.
@mekkaokereke I'm curious by what you mean by "Imposter syndrome isn't real" ?
Do you mean it's a feeling that isn't based on reality?
Or it just never exists - anyone in a psychologically-safe environment can't feel like an imposter?
(Because I read it as the second and while I wholeheartedly support your push for us to all build better environments and remove "Imposter phenomenon" I think it would be a mistake to suggests feeling are always rooted in observed reality)
@beardymcnerd @mekkaokereke I read the point as being that it's not a disorder but a rational response to knowing one will be treated in a discriminatory way for being seen as not knowing something.
So many things labeled disorders are rational responses in the same way, and framing them as a disorder is based in denial of the underlying injustice.
Impostor Phenomenon was first described by Dr. Pauline Rose Clance back in the 1970s. She described the very real feelings of extremely high performing women in academia and medicine, who were convinced that they were frauds that would soon be found out soon.
Digging into why they felt that way, revealed that many had upbringings where either:
A) They were frequently told that they were inately smart
B) They had a male sibling that they were constantly told was the smart one
@mekkaokereke Thank you!
That's encouraged me to go and read the paper. It's very "of it's time", but interesting and enlightening.
It does seem to align with what I mostly see as the modern conception: of feelings of inadequacy _despite_ receiving praise and honours in their career.
Appreciate you taking the time to reply!
@mekkaokereke I like this description, thank you.
I think I've often avoided seeing my own experience of that.
It's a limbic brain response, not a rational one. That little truth is shared with self-delusion.
mekka okereke 
Lauren Weinstein
jr conlin —〰—
jonathanpeterson
Aviva Gary
Alexander Kknli
Jeremy Kahn
Dan Hon
0xC0DEC0DE07EA
Remco van Bree
birdpoof
John Lusk
David Nash
Sasha Göbbels 🐿️
chiasm
François 🇨🇦
Julia Rez
Clay Rosenthal
Nik
Dylan
Karen Bruner
Shannon Clark
Laukidh
Sablebadger
Paul Czege
Simon Holness
Cassandrich
recursive 🏳️🌈
Stanley Jones
VulcanTourist [MOVED]
Paul