7 ms·

git is even worse than whatever these tweets suggest Office 365 does. Git stats have been used against me on performance reviews even though there's no context

git is even worse than whatever these tweets suggest Office 365 does. Git stats have been used against me on performance reviews even though there's no context to the situation. I went from a programming role to an ops role because the ops person left the team and I was the only one left that could do the work, so my number of check-ins dropped, but my contributions that improved the ops section weren't acknowledged. Any company can create bad metrics if they want. The only solution is to either play the game, or change companies. That said, I fail to see how effective employee monitoring would be via Office 365. I don't know who would be evaluated on number of emails sent per day, unless they were some sort of support personnel. And if so, they are already being monitored by other means. Any other type of evaluation sounds absurd, ex. number of Word docs opened or created, number of spreadsheets opened, number of presentations created, etc.

syshum 6y ago
Combined with that, many employee's know how to work the metrics to make themselves look more productive than they really are. So you end up with work being done to fit the metrics, not necessarily work that is of highest need or more impactful to the organization. I have seen plenty of bad metrics lead to managers no understanding why their dept's appear to be "hitting all the numbers" but also no performing in the way they desire at the same time
- ping_pong 6y ago
  After my performance review, I talked to my boss and my boss's boss, and asked them if that's really what they wanted me to do, ie. fulfill these arbitrary metrics or if they wanted me to work on whatever made the team better. They said no, of course not, but a good engineer should be able to fulfill those metrics and make the company better. I explained the situation, that the ops person left and they haven't hired anyone to take over, so I'm left holding the bag and they refused to acknowledge it, so I quit.
  - greesil 6y ago
    As a manager myself, you have terrible managers. They should be able to acknowledge the facts of the situation and not put the blame on you. To the opposite, good work, such as taking over something critical when someone left the team should have been rewarded. Metrics are useful, but to only rely on them is what lazy or incompetent managers do.
    - blub 6y ago
      What was the point of saying all of the above? In the real world there's many bad managers and new managers are assigned to existing projects all the time. Any tool which provides more misleading metrics is bad news. Proving that the initial problem can be reduced to a people problem is useless.
      - jsgo 6y ago
        my guess is they're trying to say that the person was in the right, their managers in the wrong (and they are a manager themselves) and if they run into the situation again, they know what to look for. Also, in the event this person is beating themselves up over the situation (thinking this is the norm, thinking that maybe they made the wrong decision, etc.) it would appear it is intended to put them at ease that they did the smart thing and it isn't like this on every team.
      - greesil 6y ago
        The point is that some metrics are useful tools, and that some managers are tools. To be more specific about metrics, something that measures what your users care about, or is beneficial to your company is actually useful. Sometimes the most important metric is a zero or one, which is did you ship? Measuring commit frequency seems intrinsically useless.
  - wolco2 6y ago
    Why not just ignore the DevOp work and focus on what counts. If no one is measuring ops then it doesn't matter. Someone will get upset the issue will go above you and they will magically hire someone. If you are covering up for something (no staff) than you will look bad. Push the problem up. Make the people above you look bad instead of taking every bullet.
- dane-pgp 6y ago
  > you end up with work being done to fit the metrics, not necessarily work that is of highest need Obligatory reference to (generalised) Goodhart's Law: "When a measure becomes a target, it ceases to be a good measure." https://en.wikipedia.org/wiki/Goodhart's_law https://en.wikipedia.org/wiki/Goodhart's_law
dboreham 6y ago
> or change companies Definitely that.
boogies 6y ago
At least git is Free, transparent, auditable software, designed to track the work itself, not surreptitiously log each device you glance at a document on and MS only knows what else.
- npsimons 6y ago
  > At least git is Free, transparent, auditable software, designed to track the work itself, not surreptitiously log each device you glance at a document on and MS only knows what else. Yeah, GP's comment was completely off base. O365 is user-hostile tracking-ware by design. git is just a tool with incredibly high granularity. We don't say scalpels suck because psychopaths use them to torture people.
bjohnson225 6y ago
I used to work at a company which used GitPrime, a product which tries to use git statistics to create graphs and productivity rankings for management, but of course it lacks any context. The easiest way to shoot up the ranking was to avoid any work where you'd not be instantly coding. Never fix a bug where you need to spend time tracking down the issue, never take on the more complicated pieces of work, never use your time to help out junior team members. Instead focus on simple work and trivial refactoring. The statistics are vaguely interesting and I never had it used against me, but the old quote applies: "when a measure becomes a target, it ceases to be a good measure."
- jlj 6y ago
  I once saw something similar where a co-worker on a newly formed team wanted to keep design documents in a git repo. Those types of documents can have dozens of small changes, or in this case commits, each day. It was also painful to read and edit. Luckily our manager stopped the practice early before it took root. Another anecdote, I was once in a tech lead role and spent a lot of time pairing and guiding others on their development work. My commits were low but impact was high. Low commits counted against me during promotion review.
  - goatinaboat 6y ago
    I once saw something similar where a co-worker on a newly formed team wanted to keep design documents in a git repo. Those types of documents can have dozens of small changes, or in this case commits, each day. It was also painful to read and edit. Luckily our manager stopped the practice early before it took root. There's a thing now called GitOps, where you manage your infrastructure by tweaking things in a Git repo then Puppet/Ansible/Terraform/Saltstack/whatever runs them automatically. Easy way to get "number of commits" way, way up in a plausible way. I have a colleague who creates a Jira for every tiny thing that for most people would be just a routine part of their job. He literally spends more time fooling around in Jira than he does actually doing his real job. But he looks great in the stats...
  - me_again 6y ago
    Design docs as text files (eg Markdown) in Git doesn't actually sound so terrible to me. Counting impact by number of commits is obviously dumb.
qz2 6y ago
At least you can git squash all your slacking into one commit and merge.
- dec0dedab0de 6y ago
  or make a bunch of tiny commits and change the dates to make it seem like you've been productive.
- kreitje 6y ago
  Ha! I worked at a place where every Monday I had to count the number of commits each team member made and put them on a weekly report. Once the company got bought, my new boss in charge of all developers across all divisions asked me to not mention that again so he doesn't have to do that. Once I moved on, it sounds like ParentCo took a more active role in the development and hopefully they stopped that.
  - qz2 6y ago
    Yeah that was a shitty approach. Stuck at same sort of place. I did an order of magnitude more commits than my peers last year. I had to explain that it's because commits are cheap which makes experiments and rollbacks easy, not that I'm code jesus or something.
thesuitonym 6y ago
When I interview for a job, I ask what their KPIs are. If they have an answer at all, I withdraw my application.
- MisterKent 6y ago
  Why?
- maxehmookau 6y ago
  Half of your post was reasonable.
  - chris_wot 6y ago
    The second half.
- corytheboyd 6y ago
  That’s a bit extreme. I would ask them to explain why they chose the KPIs, because that’s where you’ll get the bullshitters to break down very obviously.
- wizzwizz4 6y ago
  What about “whether you're doing the work”, or “whether your team benefits from you being in it”?
- teekert 6y ago
  We have "Inner Source" as a kpi now. Meaning we constantly ask projects where their code is, how others can use it and where an internal community can develop. I like this kpi.
- paxys 6y ago
  So you only work for companies that don't do any kind of performance reviews or promotions?
  - user5994461 6y ago
xorcist 6y ago
Gamification has consequences. I have seen git commits change over time because someone started to look at stats. You absolutely get what you benchmark for.
vishnugupta 6y ago
The overall commits/engineer graph plotted over time was one of the metrics used to justify first round of layoffs at Uber last year. Not necessarily at the individual level but to come up with an aggregate reduction target.
closeparen 6y ago
The gospel of metrics in tech management is toxic.
- aeternum 6y ago
  If you were running engineering, what would you use instead? How would you determine salaries and bonuses? Going by completely subjective metrics is pretty unfair, everyone has some amount of bias.
  - closeparen 6y ago
    Human judgement. A manager stupid enough to be worse at evaluating engineers than metrics are would struggle to keep himself fed and warm, and is therefore not a realistic threat.
coldtea 6y ago
>That said, I fail to see how effective employee monitoring would be via Office 365. I don't know who would be evaluated on number of emails sent per day, unless they were some sort of support personnel. Nobody cares for effectiveness above a certain company size. Even if the company crashes the managers will still get their bonuses and golden parachutes...
- kortilla 6y ago
  Uh, no, bonuses and golden parachutes during a failing company go to execs when that happens. Middle and even upper management below the exec level rarely gets shit if the company is doing poorly.
_jal 6y ago
I don't want to see tools built for surveillance, although that's a lost cause at this point. But the tool doesn't force managers to use it, and the absence of a surveillance mechanism, someone who seeking an excuse will invent a metric for the purpose. The only cure I've found for toxic culture is to remove someone from it. Usually me.
PascLeRasc 6y ago
Not trying to go off-topic here, but has anyone here actually used git in the workplace? I'm 3 for 3 on jobs that don't believe in version control, and other companies in my city haven't known what it is, I'm constantly asked about having it on my resume skills section. I've been told everything from "it's not enterprise ready and there's no SLA" to "open source is communism".
- raxxorrax 6y ago
  I only used git and MS team foundation server once. Never going to use that again, you cannot imagine how slow that monster was. It had VS integration, but it was such a bad product overall. Used SVN previously for private stuff and I think it is still spread in many companies. I liked it, just all the small folders are a PITA and git is faster for larger repositories. It had far better tools than git in my opinion, TortoiseSVN. Heavy, but practical. I would always use git now, even if only locally.
- offtop5 6y ago
  What type of living hell doesn't have version control? Git is the first thing I set up when I start a new side project, every job aside from one which let me set it up, have used version control. I will say Plastic SCM sucks and no one should use it though. One of my side projects ran into massive issues with my contractor due to no one, including myself understanding how plastic works. GitHub for life
  - outworlder 6y ago
    > What type of living hell doesn't have version control? I've worked (thankfully not as an employee) for a well known BANK which did not have any kind of version control whatsoever. None. Not even folders with dates. "Where's the latest release?" "Oh, it's in John Doe's flash drive over there" I am not kidding. After a lot of pestering by my company (which was working on a consulting role), they spent big bucks on some unknown version control system used by hardly anyone else. It followed the old Source Safe's "lock" mechanism. Except you had to use their own software and navigate to each file in question (no searching). Imagine how amazing it was to work on a java project and sloooowly open each folder in the reverse DNS package convention until you got to the file you wanted. Very slowly, because all actions happened on the server. Later on they were bragging on how they were able to create a branch in less than a week. Merging was a day long affair with all code contributors present. EDIT: still, that was over a decade ago. I didn't know there were companies involved with software in any capacity that still don't have version control in 2020.
Beached 6y ago
if you use a full 0365 stack there are a lot more datapoints to use than just emails sent, or documents opened/created. you have login times to see if they are late, chat metadata to see time to response on dm, idle status to see if keyboard / mouse went inactive. meetings attended, page views on the intranet, etc. if your a full stack o365 shop, then you can essentially log any and all activity for most employees, and therefore target kpi's that you want. even it admins to a degree will have these logged. powershell sessions opened, commands executed against the stack, every single click 8n the web gui in o365, every single action in azure. they are all logged and assessable, I used the same data for security monitoring before I moved to a Google shop
agentultra 6y ago
> Git stats have been used against me on performance reviews That sounds awful. Sorry that happened to you. I've been an engineering manager, for a time, and I used Gitprime to track stats. It was never used for performance reviews or in making promotion decisions or anything. What it was used for was to manage upwards and demonstrate to the upper management how decisions affect the team, to indicate when we needed to spend on hiring and had room to onboard new people, etc. I don't see metrics as "surveillance," unless they're used that way. They can be used for good. But I think as you and others have pointed out -- it requires context and transparency: you can't just rely on metrics for reporting on a person. Tech managers: you need to bring your people up, develop their skills, and pay attention! Metrics are useful but only relationships bring context. (I eventually returned to engineering as I didn't find I enjoyed management much)
Rillen 6y ago
They might have not been any indication for you but i used it as an additional marker; When there is a team member where you feel something is off from a performance perspective and they should do the same things as other colleges and there is a hard difference between them, it helps you that your gut feeling is not that off.
Wurdan 6y ago
That's the thing, it's not the existence of the tools that's the problem, it's how they're used. Nobody railed against the manufacturers of punch clocks for their products, but these days most employers have come to realize that this is a poor proxy for measuring productivity and don't use such tools.

syshum 6y ago

Combined with that, many employee's know how to work the metrics to make themselves look more productive than they really are. So you end up with work being done to fit the metrics, not necessarily work that is of highest need or more impactful to the organization. I have seen plenty of bad metrics lead to managers no understanding why their dept's appear to be "hitting all the numbers" but also no performing in the way they desire at the same time

ping_pong 6y ago
After my performance review, I talked to my boss and my boss's boss, and asked them if that's really what they wanted me to do, ie. fulfill these arbitrary metrics or if they wanted me to work on whatever made the team better. They said no, of course not, but a good engineer should be able to fulfill those metrics and make the company better. I explained the situation, that the ops person left and they haven't hired anyone to take over, so I'm left holding the bag and they refused to acknowledge it, so I quit.
- greesil 6y ago
  As a manager myself, you have terrible managers. They should be able to acknowledge the facts of the situation and not put the blame on you. To the opposite, good work, such as taking over something critical when someone left the team should have been rewarded. Metrics are useful, but to only rely on them is what lazy or incompetent managers do.
  - blub 6y ago
    What was the point of saying all of the above? In the real world there's many bad managers and new managers are assigned to existing projects all the time. Any tool which provides more misleading metrics is bad news. Proving that the initial problem can be reduced to a people problem is useless.
    - jsgo 6y ago
      my guess is they're trying to say that the person was in the right, their managers in the wrong (and they are a manager themselves) and if they run into the situation again, they know what to look for. Also, in the event this person is beating themselves up over the situation (thinking this is the norm, thinking that maybe they made the wrong decision, etc.) it would appear it is intended to put them at ease that they did the smart thing and it isn't like this on every team.
    - greesil 6y ago
      The point is that some metrics are useful tools, and that some managers are tools. To be more specific about metrics, something that measures what your users care about, or is beneficial to your company is actually useful. Sometimes the most important metric is a zero or one, which is did you ship? Measuring commit frequency seems intrinsically useless.
- wolco2 6y ago
  Why not just ignore the DevOp work and focus on what counts. If no one is measuring ops then it doesn't matter. Someone will get upset the issue will go above you and they will magically hire someone. If you are covering up for something (no staff) than you will look bad. Push the problem up. Make the people above you look bad instead of taking every bullet.
dane-pgp 6y ago
> you end up with work being done to fit the metrics, not necessarily work that is of highest need Obligatory reference to (generalised) Goodhart's Law: "When a measure becomes a target, it ceases to be a good measure." https://en.wikipedia.org/wiki/Goodhart's_law https://en.wikipedia.org/wiki/Goodhart's_law

dboreham 6y ago

> or change companies Definitely that.

boogies 6y ago

At least git is Free, transparent, auditable software, designed to track the work itself, not surreptitiously log each device you glance at a document on and MS only knows what else.

npsimons 6y ago
> At least git is Free, transparent, auditable software, designed to track the work itself, not surreptitiously log each device you glance at a document on and MS only knows what else. Yeah, GP's comment was completely off base. O365 is user-hostile tracking-ware by design. git is just a tool with incredibly high granularity. We don't say scalpels suck because psychopaths use them to torture people.

bjohnson225 6y ago

I used to work at a company which used GitPrime, a product which tries to use git statistics to create graphs and productivity rankings for management, but of course it lacks any context. The easiest way to shoot up the ranking was to avoid any work where you'd not be instantly coding. Never fix a bug where you need to spend time tracking down the issue, never take on the more complicated pieces of work, never use your time to help out junior team members. Instead focus on simple work and trivial refactoring. The statistics are vaguely interesting and I never had it used against me, but the old quote applies: "when a measure becomes a target, it ceases to be a good measure."

jlj 6y ago
I once saw something similar where a co-worker on a newly formed team wanted to keep design documents in a git repo. Those types of documents can have dozens of small changes, or in this case commits, each day. It was also painful to read and edit. Luckily our manager stopped the practice early before it took root. Another anecdote, I was once in a tech lead role and spent a lot of time pairing and guiding others on their development work. My commits were low but impact was high. Low commits counted against me during promotion review.
- goatinaboat 6y ago
  I once saw something similar where a co-worker on a newly formed team wanted to keep design documents in a git repo. Those types of documents can have dozens of small changes, or in this case commits, each day. It was also painful to read and edit. Luckily our manager stopped the practice early before it took root. There's a thing now called GitOps, where you manage your infrastructure by tweaking things in a Git repo then Puppet/Ansible/Terraform/Saltstack/whatever runs them automatically. Easy way to get "number of commits" way, way up in a plausible way. I have a colleague who creates a Jira for every tiny thing that for most people would be just a routine part of their job. He literally spends more time fooling around in Jira than he does actually doing his real job. But he looks great in the stats...
- me_again 6y ago
  Design docs as text files (eg Markdown) in Git doesn't actually sound so terrible to me. Counting impact by number of commits is obviously dumb.

qz2 6y ago

At least you can git squash all your slacking into one commit and merge.

dec0dedab0de 6y ago
or make a bunch of tiny commits and change the dates to make it seem like you've been productive.
kreitje 6y ago
Ha! I worked at a place where every Monday I had to count the number of commits each team member made and put them on a weekly report. Once the company got bought, my new boss in charge of all developers across all divisions asked me to not mention that again so he doesn't have to do that. Once I moved on, it sounds like ParentCo took a more active role in the development and hopefully they stopped that.
- qz2 6y ago
  Yeah that was a shitty approach. Stuck at same sort of place. I did an order of magnitude more commits than my peers last year. I had to explain that it's because commits are cheap which makes experiments and rollbacks easy, not that I'm code jesus or something.

thesuitonym 6y ago

When I interview for a job, I ask what their KPIs are. If they have an answer at all, I withdraw my application.

MisterKent 6y ago
Why?
maxehmookau 6y ago
Half of your post was reasonable.
- chris_wot 6y ago
  The second half.
corytheboyd 6y ago
That’s a bit extreme. I would ask them to explain why they chose the KPIs, because that’s where you’ll get the bullshitters to break down very obviously.
wizzwizz4 6y ago
What about “whether you're doing the work”, or “whether your team benefits from you being in it”?
teekert 6y ago
We have "Inner Source" as a kpi now. Meaning we constantly ask projects where their code is, how others can use it and where an internal community can develop. I like this kpi.
paxys 6y ago
So you only work for companies that don't do any kind of performance reviews or promotions?
- user5994461 6y ago

xorcist 6y ago

Gamification has consequences. I have seen git commits change over time because someone started to look at stats. You absolutely get what you benchmark for.

vishnugupta 6y ago

The overall commits/engineer graph plotted over time was one of the metrics used to justify first round of layoffs at Uber last year. Not necessarily at the individual level but to come up with an aggregate reduction target.

closeparen 6y ago

The gospel of metrics in tech management is toxic.

aeternum 6y ago
If you were running engineering, what would you use instead? How would you determine salaries and bonuses? Going by completely subjective metrics is pretty unfair, everyone has some amount of bias.
- closeparen 6y ago
  Human judgement. A manager stupid enough to be worse at evaluating engineers than metrics are would struggle to keep himself fed and warm, and is therefore not a realistic threat.

coldtea 6y ago

>That said, I fail to see how effective employee monitoring would be via Office 365. I don't know who would be evaluated on number of emails sent per day, unless they were some sort of support personnel. Nobody cares for effectiveness above a certain company size. Even if the company crashes the managers will still get their bonuses and golden parachutes...

kortilla 6y ago
Uh, no, bonuses and golden parachutes during a failing company go to execs when that happens. Middle and even upper management below the exec level rarely gets shit if the company is doing poorly.

_jal 6y ago

I don't want to see tools built for surveillance, although that's a lost cause at this point. But the tool doesn't force managers to use it, and the absence of a surveillance mechanism, someone who seeking an excuse will invent a metric for the purpose. The only cure I've found for toxic culture is to remove someone from it. Usually me.

PascLeRasc 6y ago

Not trying to go off-topic here, but has anyone here actually used git in the workplace? I'm 3 for 3 on jobs that don't believe in version control, and other companies in my city haven't known what it is, I'm constantly asked about having it on my resume skills section. I've been told everything from "it's not enterprise ready and there's no SLA" to "open source is communism".

raxxorrax 6y ago
I only used git and MS team foundation server once. Never going to use that again, you cannot imagine how slow that monster was. It had VS integration, but it was such a bad product overall. Used SVN previously for private stuff and I think it is still spread in many companies. I liked it, just all the small folders are a PITA and git is faster for larger repositories. It had far better tools than git in my opinion, TortoiseSVN. Heavy, but practical. I would always use git now, even if only locally.
offtop5 6y ago
What type of living hell doesn't have version control? Git is the first thing I set up when I start a new side project, every job aside from one which let me set it up, have used version control. I will say Plastic SCM sucks and no one should use it though. One of my side projects ran into massive issues with my contractor due to no one, including myself understanding how plastic works. GitHub for life
- outworlder 6y ago
  > What type of living hell doesn't have version control? I've worked (thankfully not as an employee) for a well known BANK which did not have any kind of version control whatsoever. None. Not even folders with dates. "Where's the latest release?" "Oh, it's in John Doe's flash drive over there" I am not kidding. After a lot of pestering by my company (which was working on a consulting role), they spent big bucks on some unknown version control system used by hardly anyone else. It followed the old Source Safe's "lock" mechanism. Except you had to use their own software and navigate to each file in question (no searching). Imagine how amazing it was to work on a java project and sloooowly open each folder in the reverse DNS package convention until you got to the file you wanted. Very slowly, because all actions happened on the server. Later on they were bragging on how they were able to create a branch in less than a week. Merging was a day long affair with all code contributors present. EDIT: still, that was over a decade ago. I didn't know there were companies involved with software in any capacity that still don't have version control in 2020.

Beached 6y ago

if you use a full 0365 stack there are a lot more datapoints to use than just emails sent, or documents opened/created. you have login times to see if they are late, chat metadata to see time to response on dm, idle status to see if keyboard / mouse went inactive. meetings attended, page views on the intranet, etc. if your a full stack o365 shop, then you can essentially log any and all activity for most employees, and therefore target kpi's that you want. even it admins to a degree will have these logged. powershell sessions opened, commands executed against the stack, every single click 8n the web gui in o365, every single action in azure. they are all logged and assessable, I used the same data for security monitoring before I moved to a Google shop

agentultra 6y ago

> Git stats have been used against me on performance reviews That sounds awful. Sorry that happened to you. I've been an engineering manager, for a time, and I used Gitprime to track stats. It was never used for performance reviews or in making promotion decisions or anything. What it was used for was to manage upwards and demonstrate to the upper management how decisions affect the team, to indicate when we needed to spend on hiring and had room to onboard new people, etc. I don't see metrics as "surveillance," unless they're used that way. They can be used for good. But I think as you and others have pointed out -- it requires context and transparency: you can't just rely on metrics for reporting on a person. Tech managers: you need to bring your people up, develop their skills, and pay attention! Metrics are useful but only relationships bring context. (I eventually returned to engineering as I didn't find I enjoyed management much)

Rillen 6y ago

They might have not been any indication for you but i used it as an additional marker; When there is a team member where you feel something is off from a performance perspective and they should do the same things as other colleges and there is a hard difference between them, it helps you that your gut feeling is not that off.

Wurdan 6y ago

That's the thing, it's not the existence of the tools that's the problem, it's how they're used. Nobody railed against the manufacturers of punch clocks for their products, but these days most employers have come to realize that this is a poor proxy for measuring productivity and don't use such tools.