21 ms·
> I drastically reduced the time developers spent repairing those failures, but there were no metrics that tracked developer time. For several jobs in a row, I
by dahart 9y ago
> I drastically reduced the time developers spent repairing those failures, but there were no metrics that tracked developer time.
For several jobs in a row, I've felt that helping others on a team is undervalued and under-recorded. I've been planning to implement the "assist" metric, similar to basketball, on my own team for a while.
The idea would be something along the lines of everyone gets a set of assist points they must distribute to people who help them the most.
While I don't love the idea of gamifying it, all the places I've ever worked have had too strong of a reverse incentive towards individual achievement, especially when it comes to promotion, and not nearly enough support for teamwork in specific and measurable ways.
Anybody have experience with teamwork metrics like this, or others that worked or didn't work?
- xexers 9y ago> I've been planning to implement the "assist" metric, Anyone from Shopify here? If I recall correctly, this was tried at shopify and has since been removed due to people gaming the system for their friends.
- AndrewKemendo 9y agoThe idea would be something along the lines of everyone gets a set of assist points they must distribute to people who help them the most. There is a really great idea in there somewhere - create incentives for helping others - but I'm struggling to come up with a good way to implement it that doesn't just water everything down. The closest I can think of is how repo commits can be counted/graphed per user. Works easily for code jockeying, but less so for softer talents like pair programming.
- sharemywin 9y agoWhat most companies do is give bullshit senior level promotions that way they can get 5 good years out of someone before they get the same realization it took him to get to in 2 years.
- bluetwo 9y agoLloyd Shapley already figured out a "fair" way to do this, called the Shapley value. Basically it's a thought experiment in which all team members of a project (or group, etc) estimate the impact when losing various members of the team (whose work presumably is done by other team members or contracts) on the final cost. Run some math on the data and you have a group estimation of the relative contribution of all members. I have a theory that the reason more managers don't use this available tool is that it would probably rate management contribution on the lower end...
- sharemywin 9y agoThat's what's dysfunctional about a lot of companies/our society. You persuade others to do/[help with] your work you are showing leadership and get promoted. Sorry, you got duped into having sympathy and showing teamwork. The better you are at convincing others we're a team and hard work is it's own reward when it's really just dog eat dog the higher up the food chain you go.
- stephencanon 9y agoI've never tried it, but I suspect you'll run into the same issue that sports does, where "assists" don't accurately reflect the people who are making things happen on the court. We miss the perimeter shooter who positions themselves just so to pull two defenders out of position, but give credit to the point guard who makes the now-easy pass for the "assist". It's entirely possible for the person most responsible for a play succeeding to never touch the ball. You really want something like value-above-replacement, but that's pretty hard to measure without a clear definition of what "success" is and reams of data.
- bob_theslob646 9y agoYou're forgetting that you could also track secondhand assist as well. I can't speak for programming but in soccer this is well-known. It's just very hard to Value just like an employer would in terms of having that person on board for morale. Fortunately the better people get it and understand that that person should be value even if they weren't directly involved in making the change but they were definitely a part of it. This is what plagues me with most corporate companies in the United States. It should be about teamwork not about what this person did or I did, rather what the team did. Sure you can value people individually, but teams trump individuals all day long in the long run.
- ajmurmann 9y agoThat's a huge reason why I strongly dislike OKRs especially in an Agile environment and even more when people pair program.
- freehunter 9y agoA person’s projected morale is very under-appreciated. I don’t know if I’m at the point in my career that I boost morale just by being on the team, but I do know there are people who, if they left my company, would have an impact on my morale even though I never ask them for help or work directly with them. Just something about knowing a person is an amazing and knowledgeable resource and he’s choosing to work here of all places is a subtle morale boost. When they decide to leave, others might start asking “well if that guy left, am I still making the right decision to stay?” I’m sure we’ve all experienced someone important leaving our company and having a waterfall of turnover that follows it. That is very disruptive and more should have been done to keep that one employee or at least keep the rest that followed them out the door.
- shados 9y agoIn one of my previous teams, we did away with individual metrics because they "punished" team players. If an engineer spent most of a week helping another with their issue, it made them look bad in whatever tracker we used by almost every metric. I prefered sticking to trying to optimize whole team velocity and using 1 on 1s and peer reviews to track individual performance where it was needed (eg: compensation review and promotions). It's still not perfect: people could be spending them helping someone from another team and that would go under the radar (except the peer reviews), but it still worked a lot better.
- Isamu 9y agoYeah we don't do metrics, we gather short appraisals from a list of people that you provide, that you worked with throughout the year. You are encouraged to provide names of people that are outside of your group. So if you helped some other team, this is the way you cash in on that. Helping other teams succeed is a cultural thing and your organization has to encourage it with its practices and procedures.
- babesh 9y agoThat is gamed too. You just don't include the names of people you have wronged: the bugs you have left for others to fix, the undocumented code, the projects you were supposed to complete but got reassigned so that you could work on something with higher promotion possibilities, the delays you cause in other people's work, etc...
- longerthoughts 9y agoThis is exactly what I've seen happen in my experience and have often wondered about alternatives. Bridgewater, for example, uses a system that allows for feedback from anybody at any time [0], but I suspect that presents new issues like negative reviews from rival employees. Additionally, a lack of context could result in negative feedback for decisions that actually help the business (e.g. passing on requests to help colleagues with relatively low-value items in favor of focusing on higher priorities). [0] http://www.businessinsider.com/bridgewater-employees-rate-performance-dots-2017-2 http://www.businessinsider.com/bridgewater-employees-rate-pe...
- redux13 9y agoThe company that I work for employs a system of peer review - where at the end of the year your teammates write reviews for you and your appraisal is based on that. It seems to work pretty well our case and definitely aligns the incentives of the employees with having an impact on the people around them.
- greedo 9y agoThat's assuming that your peers aren't willing to sabotage you to get ahead...
- tankenmate 9y agoIf you are going to hand out points for people to vote with, may I suggest quadratic voting.[0] [0] http://ericposner.com/quadratic-voting/ http://ericposner.com/quadratic-voting/
- Djvacto 9y agoSo, would you mind helping me see if I understand quadratic voting correctly? Let's say there are 4 voters: A, B, C, D, voting on policies 1-10. Everyone gets a certain number of votes (say one per policy, so 10 total). Voter A only really cares about policy 5 & 7. So they want to buy votes for these 2. They can give up their votes in, say, policies 1-4 to get 2 more votes for policy 5, and can also give up votes 6, 8-10 to get 2 more votes for policy 7. So they put 8 votes into the "common pool" in order to ensure that, in addition to their vote for 5 & 7, each policy also gets 2 extra votes?
- chrisabrams 9y agoI think you just described a typical day in Congress...
- winstonewert 9y agoIn QV, you buy votes with money. You can buy as money votes as you want, but you have the pay v*v for it, where v is the number of votes.
- dahart 9y ago> In QV, you buy votes with money. FWIW, the paper describes it as a vote currency that can only be used for the vote, as opposed to cash you could spend on anything. The voice credits have to be distributed to voters. 'individuals pay for as many votes as they wish using a number of "voice credits" quadratic in the votes they buy.'
- winstonewert 9y agoIf you are referring to the same QV paper I read, it spends most of the paper arguing for money but then at the end suggest a virtual currency at the end for those who can't accept their argument for allowing buying votes with money.
- aecorredor 9y agoAt my company we use something called bonusly. You get 30 bonusly per month to give out to your teammates, and each bonusly is worth $1. You can buy basically anything with the bonusly you receive (amazon gift cards = anything). On top of being able to buy stuff, it also keeps metrics on which teammates help the most. The only downside is that you must give out all your bonusly at the end of the month (if not it's wasted money), so sometimes you start giving it out without really taking into consideration which teammate has helped you the most.
- Jemmeh 9y agoIn my experience this often just becomes a popularity contest/give to whoever you eat lunch with. I can have my head down all week neck deep in code but because of this I'm not interacting with people around me as much and basically get punished for focusing on work.
- cookiecaper 9y agoYes. You can't change people. Your actual job description is worth 20% or less of your workplace success; as long as you're doing enough to stave off disaster, you are pretty much fine. There is a simple metric for workplace success: if most people are happier while you're around (including and especially anyone above you in the chain of command), you're winning; if not, you're not. This can probably be summarized as "popularity". A few things can overcome this, so you may want to be careful not to go 100% focused on it, but you will definitely prosper if you make it the focus of 95% of your work. Developers who like to debate and don't take it personally when a colleague disagrees are at a distinct disadvantage here, despite simultaneously frequently being some of the best qualified coders, because they don't understand that they're rare, and that many other people take any form of non-ambiguous criticism of their ideas or work as a major offense: taking bread out of their kids' mouths, etc. I have a theory that the more a person cares for niceties, appearances, and superficialities, the more it shows that they're incapable of operating at a level any deeper than that (that is, capacity and the presence of superficial tokens generally understood to mark capacity are inversely correlated; e.g. an expensive watch does not necessarily indicate power, wealth, or importance; compare Dunning-Kruger Effect). People who are capable of doing useful work seem to find these superficialities asinine and have difficulty hiding it, whereas people who aren't capable of useful work try as hard as possible to keep the focus on the surface level, because that's their only hope. Since the surface level is immediately apparent and the importance of the subsurface may only become apparent after some experience, imposters can, and frequently do, find great success.
- dep_b 9y agoI worked in a remote team once helping a company in Seattle (that perhaps had a bit of a jock culture problem). I could be stuck for days because following the README in a repo just wasn't enough to get the project to compile and run. Every standup I was telling "I'm completely blocked for the last two days because I cannot run the repo, so I cannot run any tests and at the moment I'm doing absolutely nothing.". To be sure to repeat a message like that every four hours in channels like Slack. Managers read it, my teammates in Seattle read it. Nobody ever helped us (because you know pair programming is so much more efficient if you can have two programmers blocked when you could just have one staring out of the window and typing "help" to invisible people). If at all I would be unblocked after a few days usually the problems was some kind of unguessable piece of configuration that should have been part of the README and/or code, a problem completely unsolvable without knowing what to do. Only then I could continue doing a minor change like updating a CSS file for the next Safari or IE quirk, usually just 5 minutes of work. Apparently helping remote teammates to get unstuck in their work just wasn't an interesting performance metric. Until this day I still can't believe that everybody was just throwing tens of thousands of dollars of work out the window because helping stuck teammates wasn't counting towards some kind of bonus.
- KKKKkkkk1 9y agoWhat makes my stomach turn is the interpersonal aspect. How can all your teammates see you fail through no fault of your own for days and pretend they don't notice. Unfortunately in my experience, that's not a rare occurrence.
- derekp7 9y agoI can give some insight from my own experience. In my current position, there are a large number of "things" that I've had to figure out, either by reverse engineering, reading code, or intuition. So I get constantly bombarded with helping others get unstuck. Well, when it comes time for a project meeting, those on the meeting are far removed from the other groups that I constantly have to help out. And it sounds like I'm just giving excuses of why my work isn't done. So if there is a micro-crunch coming up (i.e., I need to have an SOP drafted, tested, etc by this afternoon), the entire morning I tend to blow off other peoples issues. Until I can get a chance to come up for air. But that makes me feel bad, so what I end up doing instead is I go home at the end of the day and work till past midnight just to get caught up on what I couldn't do in the daytime. For any outside-the-box work I tend to pull 20-30 hours stints over the weekend. After 25 years of this, and as I need more sleep since I'm getting older, I can be a bit snippy when someone breaks me out of my concentration because they can't figure out something, and they have a down customer that needs service "right now".
- somberi 9y agoI am manager who has a few levels under me (Total team of about 100 people). I find it difficult to buy into your experience, and hear me out. I am arguing in your favor. If you are assisting, and this includes providing support for someone else's code, or doing "invisible" tasks (setting up a redis cluster or writing end-to-end tests), this often finds its way into one email group or another. In fact I remember the names of the "helpers" as you put it by the virtue of their emails and to that end I value them, just as much if not more, for keeping us shipshape. IMHO, you should find these emails, and I am sure they exist already, and use it as your metric. Scott Fitzgerald wrote “Show me a hero and I’ll write you a tragedy.” I say show me email chains, and I will show a hero. Of course I am not stating the obvious fact - which is that your manager, if he or she is worth their salt, should totally value the role "helping" plays in making good software.
- taurath 9y agoI've spent a few weeks working on an extremely high priority business issue and the only artifacts are a few issue tracker stories, my team's slack channel, and a "huge thanks to @...." on a public chat. If you're at a high enough level that you can't value someone's contribution from those artifacts, then you need to be trusting those under you and asking about people... you'll be missing why high performers are leaving.
- wehadfun 9y agoYea if a developer helps someone there is no guarantee that there will be an email chain generated as proof. If you truly believe this in all due respect you are out of touch with reality. People still walk up and ask for help. Maybe not in remote teams but in the office all the time. Supposedly making these type of interactions possible is the whole point of open offices
- paganel 9y ago> People still walk up and ask for help In my experience people also ask for help through private messenger discussions (slack/skype etc), which of course don't leave any "official" trail and are not supposed to. Of course, a not-so-bright manager could force his/her direct reports to only use "official" channels for work-related communications, but then people would be afraid/ashamed to ask "stupid questions" in public/through an official channel and then disaster happens (it's also my experience that lots of serious bugs have been caused by people afraid/ashamed to "just ask").
- exelius 9y agoNope; the benefit of being a good team player is that it helps you build relationships. Those relationships often turn into opportunities down the line. But it’s hard to drive that behavior at an individual level without creating some conflicts of interest. This is why I generally dislike measuring individuals against specific metrics. I try to set individualized metrics for all of my direct reports — for example, if I had a developer who keeps checking in code without valid tests, code coverage might be a metric I track with him just to help him understand how he’s doing against his personal goal. Even within the same job description there is often room for a significant variation in technical and interpersonal skills, and you need a team with a good mix of both to be successful. A one-size-fits-all approach to individual metrics just leads to a team of people with the exact same skill set. I think the way you measure teamwork is at the team level by having a consisitent set of team-level metrics the entire team is judged against. This means individual personality and career path variations don’t matter as much. I’m talking things like say, number of broken builds, velocity improvements, production issue counts, etc. These should be business outcomes that the product folks care about. If you make it the team’s responsibility, you’ll find teamwork happens more naturally. This approach to performance management is in direct opposition to stack ranking, and a big part of why I refuse to work for a company that forces its managers to stack rank within teams. The question of “what makes a good employee great?” is so difficult to answer that if you have multiple high performers, it’s effectively arbitrary. It also tends to punish people who are maybe less technically brilliant, but have strong interpersonal skills (after all, how else does a mediocre programmer make it through a CS degree?) Those interpersonal skills often lead to better technical discussions and avoidance of miscommunication between technical folks.
- eeZah7Ux 9y ago> the benefit of being a good team player is that it helps you build relationships. Those relationships often turn into opportunities down the line. Most of the time other people leave or move to other teams. Being a decent human being is not about getting practical advantages - otherwise sociopaths would be the most helpful people around. Instead, many large corporations want to instill a culture of competitiveness and often cynicism because competing is what they do. Cooperation, lasting relationship between employees, mutual help, solidarity are increasingly scary words for many companies.
- bryanrasmussen 9y agomaybe I'm being optimistic but a couple days ago I spent basically the whole day helping another developer getting a fix done so they could make a pull request before the deadline, I think that was noticed because we are in an open layout so me telling him do first A, then B, if you do C that won't work. It isn't working in QA because of this reason etc. etc So anyway I think that was probably valued by our project manager who is also a former techie, because she probably realized if I didn't do it, that PR wasn't going in that release.
- dwaltrip 9y agoThe first concern I had is that assist points that people assign to other won't be directly comparable -- so an "assist leaderboard" might be misleading and potentially cause more trouble than it's worth. A point that Dev 1 gives to Dev 2 will not likely be equal to a point that Dev 3 gives to Dev 4. This will be especially true if everyone has the same quota of assist points to assign, as people may very easily receive different amounts of help, and thus their "actual help per assist points assigned" values will differ.
- akvadrako 9y agoI don't have much experience with this, but I find it a very important topic. My opinion is that there are so many different aspects to teamwork and social cognitive biases are so huge, it's an intractable problem to do on an individual level. The only thing we know matters is the team does well or not. So I suggest trying different methods of building/running a team, and let them go. See which teams succeed and try to copy that success or give them more power. To give some flavor to my idea - maybe the level of helpfulness you should ideally give is very low. Perhaps it enables weak members to suck up lots of time. We just don't have good data in this area.
- lhorie 9y agoIn a previous company, there was a feature in the intranet to give kudos to other people. Everyone in the company could see a kudo as a post in the global intranet "chat" sidebar (not really chat, more like a communal facebook wall of sorts that would bump posts to the top any time anyone commented on them) You could give as many kudos as you wanted, and it had zero impact on performance appraisal/raises, but it sure gave people the warm and fuzzies reading about random mini-stories of someone going above and beyond here and there. I think it was pretty effective at fostering a culture of collaboration, and it also helped surface issues-masked-as-heroics, such as people working chronically late on some projects. It also gave greater visibility to the existence and importance of the various departments, which one might not have gotten an appreciation of otherwise.
- ChuckMcM 9y agoI think I may have seen a system like that, it had a sort of bluish tint to it :-). Its interesting that throughout my career I've found it has been challenging to communicate group and individual contribution in meaningful ways to people without the whole picture. That is especially true when the activities don't translate directly to sales revenue. As a result, you get people who leave who were much more important to the organization than the organization understood and the folks left behind are stuck trying to recover and make up the loss. If there was a manager measure of 'OQ' (organizational quotient) it would measure the ability to recognize and optimize the net productive energy of an organization. Groups without that awareness seem to wither as they pursue one or two 10x employees without regard to how to support their activities.
- mpeg 9y agoAt SFDC we bought and used internally a tool called Rypple (later work.com) that was precisely that. As you say, zero impact to people in terms of their performance reviews (which were a topic of discussion in their own, each employee having to write a mini-manifesto for their year ahead) and kinda became a popularity contest where people were giving kudos to their office buddies. And don't get me started on the people that left heart stickers with messages like "thank you for starting the best company in the world" on Benioff's profile.
- hinkley 9y agoEvery lead developer position I ever got came from helping people, but that doesn’t always stop the judging on output. On one project I shaved about 40 minutes to an hour per day of finger twiddling frustration off of everyone’s work. For the team size that paid for itself in under two months. Only ever got appreciation from subordinates and flack from my good for nothing boss. But what I figured out not long after that was that you understand the work a lot better when you see it from other peoples perspectives, and you only learn about that if you are willing to help and to be candid about the state of the project. People open up. They share maybe-bugs before filing them (especially from the QA team) and then when the inevitable OH SHIT meeting happens you are the only one who has had time to think about a solution ahead of time.
- michaelgreen 9y agoThe first thing that comes to mind is peer reviews. Although it's not inherently quantitative data and doesn't remove any of the bias/politics/etc. it can be a good indicator of who others feel are helping them. However as the writer of the article pointed out often times you can't see when other people are improving your ability to do work whether that's through writing more documentation, improving some internal infrastructure, or other often invisible things that can drastically improve your life as an engineer. I think an effective approach would be through time tracking (i.e: writeup on how you spent your time for the day) in conjunction with some analysis on how the time you spend on certain things correlates with what other engineers are doing. This of course would rely on everyone being detailed in their writeups and the analysis still might not catch certain things because it's difficult to write down exactly how you spent your time. Perhaps you could only write down the three most difficult/annoying problems you ran into and just track that. I remember reading about some company that is trying to make this easier by tracking computer use and creating graphs of how people interact within their organization and how that changes over time but that doesn't seem like it would work everywhere. I don't know if there is a way to truly quantify teamwork or organizational impact. I think the important thing is to encourage your organization to communicate more: talk about what you're doing, why you're doing it, and how it helps others within your organization. If your organization is fairly big write about it and share your writeup internally. Train people within your organization to speak up if they feel they aren't being appreciated for the impact they made. Ultimately I think it's an extremely hard problem because it stems from the question of what having a positive impact means. The answer to that can branch off into millions of different small tasks that could be totally unique to a subset of people within your org. ---- The best you can do is: * Measure what you can. * Know that's not the entire picture and it never will be. After that try to create a culture that: 1. Speaks up about their own impact 2. Shares the impact of others' work, especially of those who might not speak out much 3. Is openly appreciative of and encourages 1 and 2.
- googlethrow 9y agoGoogler here: The google performance and promotion process does, in fact, rely on reviews from your peers. The process happens once every 6 months, every other being non-mandatory. Although your peers saying you were helpful is not enough, you need to demonstrate impact (and other attributes depending on your level).
- trashcan 9y agoThe company I work for has a kudos system. You can give any gift, but in practice everyone gives $25 amazon gift cards since they are almost cash. Employees are encouraged to give at least one a month to someone who deserves it. I haven't been through a performance review yet, but I wouldn't be surprised if they are taken into consideration. From my experience it works amazingly well. It's silly, but if you are ever having self-doubt, a gift card from a coworker really helps.
- xfibre 9y ago> For several jobs in a row, I've felt that helping others on a team is undervalued and under-recorded. I completely agree with you on this. Also, I like the idea of formalizing with "assist" metric, so that people are sure that their work won't go unrecognized.
- Jemmeh 9y ago99% chance of just being "friend" points instead of having anything to do with code.
- runevault 9y agoOnce a team is settled in this can be hard to measure. For me personally I was able to show mine by changing teams and helping drastically inflate overall team velocity, not just by the number of points I added, but how much more productive everyone else became. It is hard to do on an established team where you're already part of it, though, and that's a problem I don't know how to solve.
- icebraining 9y agoGo on vacation for a decent period of time? In finance, key employees have mandatory two-week leaves, which serve to detect embezzlement. Here we could have the same, but to detect positive influences.
- runevault 9y agoThat's an interesting point. Once you've been with my current employer long enough, you actually get a 1 month sabbatical which may work for that.
- muthuraj57 9y agoIn our team, we have our custom implementation of a web app like https://bonus.ly https://bonus.ly . The idea is, every member would be given a certain points(we call it jems) at the start of the month. He/she can give those points to other members who help them with the reason specified and it will be posted on a slack channel through slackbot. Using this, everyone can know what other people is working and who is helping who like that. For every month at a team meeting, people who got more points will be appreciated in front of the whole team. The total points is calculated using a formula such that a member who just start giving/receiving more points per jem than other so that new comers will be appreciated more. It works really well and at performance review, managers will consider this stats. Also the performance review contains two peers (chosen by us), and all kind of managers who we are reporting.
- zhausjgzga 9y agoAt my job, we use HeyTaco, a Slack bot. Every day you can give up to five tacos out by @mentioning somebody with a :taco: emoticon. If you answer a question or help somebody out, they will probably give you some tacos. The app keeps track of who gets the most tacos with leaderboards and the like. This has instilled a culture of helping others out on the team, by gamifying it and making it visible.
- tomkarlo 9y agoWe use something like this in my org, and Google has "kudos", but I've found that the problem is their usage is highly variable even within a medium-sized or (we're roughly 1,000.) On some teams, the average employee will have hundreds of these kudos, whereas on others, the average might be 1 or 2. This was highlighted to me recently when someone in Slack noted that one person had ~50X the kudos of everyone else and asked how that happens... and they said "just go hang around in team X's channel, they'll give you a 'taco' any time your answer a query." Given Google's promo system (where a committee looks at candidates from very different teams) this then means they'd have to try to normalize the data for it to be useful.
- r00fus 9y agoCould this be combined with a group-level kudos/tacos metric? Even if it is - don't be that manager of the group with the least group kudos/tacos?
- tomkarlo 9y agoSome roles (ops, sales) are just naturally a lot more inclined to use these kinds of systems. I generalize like that after having watched similar systems at 3 companies; you hardly ever see an engineer give a "taco" for a code review, but someone in another team might give one simply for someone knowing which room a meeting is happening in. It's hard to normalize the culture. I think peer bonuses (which usually require manager approval on one or both sides) tend to be a bit more normalized across teams.
- blaketaco 9y ago
- Stratoscope 9y agoI don't have an answer to your question, but I've certainly felt that pain. Years ago I worked at a desktop software company that hired me to be their "Windows expert". They had a team of strong programmers who mostly had a Mac background, and they wanted to improve their Windows offerings. I often spent a couple of hours a day doing what I called "house calls": finding out who was having trouble with a Windows issue and stopping by to lend them a hand. This was before "pair programming" was a common term, and full time pair programming would drive me nuts, but these house calls were very productive. The other team members knew the product code much better than I did (being a newbie), and my Windows experience helped them work through platform issues. Plus sometimes just having another pair of eyes look at a problem. Often a 20-30 minute house call would save a team member many hours of work. At one point, a teammate thanked me for helping with a tough problem and asked me, "How can you afford to be so generous with your time?" Some time after that, the VP of engineering (who I'd worked with previously and was on good terms with) asked me if everything was OK. He had noticed that the other developers on the team had lately been more productive than ever, but my own productivity wasn't quite keeping up. I think I mentioned that some of my time had gone into helping the other developers work through their Windows and other issues, but I didn't have the presence of mind to say, "Maybe my help has been one reason their productivity has gone up. I thought that's what you hired me for. Can we make it part of my job description?"
- rwitoff 9y agoYep - See some of the work my last team published here while we were rebuilding deployment tooling: https://blog.coinbase.com/scaling-developer-productivity-d23ce491f869 https://blog.coinbase.com/scaling-developer-productivity-d23.... For hard to measure metrics, softer measurements like repeated team surveys work fine too. I'd be cautious of directly measuring "assistance" outside of peer input in perf to avoid an unhealthy incentive. The most helpful people I've worked with in the past tend to grow through strong peer reviews and having the most opportunities to join new projects. The OP here really misses the point of demonstrating impact. Doing the right thing ethically then for the business is a strategy that is, well, rarely the wrong thing. Optimizing just for getting promoted might is a greedy strategy that might get you promoted once, but good luck finding peers that want to work with you again.
- paulcole 9y ago>The idea would be something along the lines of everyone gets a set of assist points they must distribute to people who help them the most. Just talk like normal people. The more you talk openly about who's contributing and who's not the easier it becomes.
- notjustanymike 9y agoWe use bonusly. As a proof-of-concept, I imagine it'd tackle some of your requirements. Bonusly points are tied to real world rewards, so there's incentive to use it. People who had a lot of "assists" at our company regularly brought in 1000+ points a month, which then translated in $100+ in amazon gift cards or steam points.
- josephjrobison 9y agoI academia, if you're a PhD student that writes the paper, you're the "first author", but anyone who helps you do research or process data or other contributions is a "second author". I think that's a better system, because one person may have 2 first author publications and 3 second author publications. Another person may have just 1 first author publication but 12 second author publications. So you can weigh and evaluate each person based on their assistance and contributions to others.
- __bjoernd 9y agoThis system is actually designed to disincentivize assistance because only first author publications will net you scientific credits when it comes to jobs or funding.
- ma2rten 9y agoGoogle actually kind of has a partial solution for this. Your performance review is not just based on manager's assessment, but you also need to gather reviews from your peers. However, you need to strike a balance between helping out peers and having impactful projects.
- pravula 9y agoOne previous company had a system where one can reward someone who assists you, in cash. It shows up on their w2, post tax and in performance reviews. The amount you can reward depends on your level.
- batmansmk 9y agoWe estimate and log our time at my company Matters. Teams estimate a large budget if they don't know how to do it. When you assist another team you generally deal with the task in way less time than planned, therefore performing greatly. It helped us discovering the importance of assists, so now we have support teams dedicated in assisting project teams. Project-driven environment + no estimate or log = politics
- mlazos 9y agoI work at Microsoft, and I know around promo time in addition to people on your team, managers ask for feedback from people not on your team that you’ve interacted with as well so that can be used in conjunction with whatever projects work you’ve completed.
- cryptbe 9y agoAt Google you can ask your coworkers to write feedback that would be included in your promo package. I wonder why the author didn't ask the developers that they helped to write a few sentences explaining how much time they'd saved thanks to the author's help. The hiring committee wouldn't ask for any metric if there were already a few people testifying that "I saved XX hours". Disclaimer: I work at Google.
- kurtisc 9y ago>XX hours Isn't that a metric?
- CodeMage 9y ago> For several jobs in a row, I've felt that helping others on a team is undervalued and under-recorded. I've been planning to implement the "assist" metric, similar to basketball, on my own team for a while. The metric might be a good idea, but I think the real problem goes deeper than that. Take a step back and ask yourself a more fundamental question: should everyone who does good work be promoted? By definition, promotions are limited and each level is more difficult to attain than the last one. Obviously, you can't promote everyone. Even worse, not everyone will be happier when promoted. So why does literally everyone want to be promoted? Why do they work so hard to get it? It's because there are no other ways of rewarding good work. And by "rewarding", I don't only mean "giving people recognition" or "giving people fun work". Those are important and should not be overlooked. Nevertheless, if you don't also give people money, sooner or later they'll fall behind their needs or just the rising costs of everything around them. Then they'll have to choose whether they need money more than intrinsic rewards. And then we're back to the original dilemma: do I play the promotion game or do I just bugger off somewhere where I can get what I need? TL;DR: Retention shouldn't be only about promotion. Give people real rewards and incentives for doing a good job at their level.
- lutorm 9y agoBy definition, promotions are limited That's not really the case if the promotion ladder is a skills/experience based ladder and not a responsibility ladder. If being a "senior engineer" just means you have reached a certain level of proficiency, then there's no reason everyone can't be promoted to that level.
- Torwald 9y agoIf you gamify you also have to maintain the game. Players will game the system, the mere fact that the game exists itself will change the overall dynamics, this and that everything is in constant, natural change, all makes it necessary to maintain the gaming rules.
- lkrubner 9y agoOne easy way to track this is tell every employee that each month they get 3 upvotes and 3 downvotes. The votes are anonymous. (If you have less than 20 employees, then maybe just 1 upvote and 1 downvote.) They vote for their co-workers. You'll see patterns in the upvotes and downvotes. The upvotes might signal "This person helped me a lot." The downvotes can catch interpersonal problems that might otherwise be hidden from you. For instance, someone on your team might be guilty of sexual harassment, but they are clever about hiding it from you. Your only clue will be that, say for instance, all the women keep downvoting that person.
- amenod 9y agoIf you can see this from the data, then the votes are not anonymous. They are just not published to other members of the team. While I appreciate the goal, I have been put in such situations before and must say I hate such questionaires and ratings. Of course they are not anonymous. And if that is so, do I tell the truth or do I lie? I don't want to lie. But I also don't want to cause trouble for my coworkers because of some misunderstanding. Admittedly, I don't have a better alternative though.
- dmayle 9y agoYou can still collect this data anonymously. Rather than giving people a fixed number of votes, you have them vote for everyone, but use random methods to conceal the source. (This has the effect of hiding single complaints, though) For example, when downvoting: flip a coin, if it comes up heads, downvote that person. If it comes up tails, only downvote that person if you think they deserve to be downvoted. Doing it this way, each person can be expected to get roughly 50% of people downvoting them. If the number is enough higher than that, then that person probably deserves the downvotes, but you can't tell who they came from.
- mkirklions 9y agoI dont work on software, but shouldnt there be deliverables that matter? Track management only.
- epicureanideal 9y agoIt would be nice to have some sort of metric for this. I haven't experienced any. As a former manager, the closest I came was making a point to observe who was helping who within the team. And following up by asking people during their 1 on 1s if they felt anyone was being especially helpful to them, explained things well or not, etc. People who helped their teammates very much got a boost in their prestige in my view, and that was good for additional job security or additional pay. If they struggled with a couple tasks, but had helped some teammates when they DID understand what to do, that very much increased my willingness to help them or give them extra time to figure things out.
- latj 9y ago"I've been planning to implement the "assist" metric, similar to basketball, on my own team for a while." Sounds like a super fun place to work. ;)
- Itaxpica 9y agoThe irony is that Google actually has a separate eng job ladder (Software Engineer, Tools and Infrastructure) which is designed to reward that kind of work specifically. If he had ladder-transferred (a fairly easy process) and gone up before a SETI committee with that same data (peer feedback on stuff like saved development time carries weight that metrics may miss), it probably would have turned out very differently.
- etse 9y agoTransferring ladders seems non-trivial. If you know ahead of time that you are doing "the dirty work", which usually you don't, do you know how long you are going to do it for, and if it makes sense to apply for a transfer? Would your manager approve of your ladder transfer if he/she needs to back-fill your position when a good engineer is hard to find or keep to begin with? (Actually, there are scenarios where a manager wants to keep underperformers from transferring out, too) Lastly, did you really want to be a Tools and Infrastructure engineer to begin with? It started with just being scrappy in order to get the project to a place where you can contribute shiny new features, which everyone is waiting for you to do, yourself included. If you transfer ladders, what are the chances you will be able transfer back in any reasonable amount of time to get credit for doing a good job on what you wanted to work on? Michael's work situation seems prevalent through a Google-like org, and a ladder transfer is unlikely a realistic solution for that.
- YZF 9y agoWhere I work we are big on peer review. So in theory helping others should have value. It's still not perfect. For example, a junior may not be able to asses the value of your help, or anything else for that matter. People will still tend to rank their friends higher. While your boss doesn't rank you whether he promotes your work to others has a big influence on how they view you. In a previous company we had a better mix of peer reviews and input from technical leads (that were not your bosses). This was by far the best system I've seen to date for a company of that size. In small companies there is typically good visibility into what everyone is doing. If you have good management then things tend to work very well. If you have bad management it can be worse than large companies. One thing I've come to learn is that this sort of stuff, call it culture, has to be a priority for the company's leadership, they have to keep inspecting it and fixing it. They need to talk to people and understand their concerns. Otherwise it always drifts the wrong way. The Google system from what I read here sounds terrible. EDIT: Having people try to optimize to play the system, just like the article describes, is inevitable so you must ensure the incentives are aligned. Make sure you find ways to look at the stuff that really matters and make sure you're not antagonizing your employees with your review system.
- nojvek 9y agoAt my workplace (Mixpanel), we use youearnedit.com. Everyone gets 2000 points per quarter and you distribute them out to whoever helped make you life easier or reward others for their effort. It has slack integration so sending /yei @someone is super easy. I probably do it multiple times a day. I have to say, after it was introduced, there definitely was a slight shift in culture. I strongly believe your 10x programmers are the ones that help the whole team become 10x by building great tools and showing best practices that makes everyone move faster.