I know the policy doesn't really exist anymore, but I still like the idea a lot in principle (for large and financially secure startups, say 1k developers and up). Just wondering what Xooglers and others think about it.
I've been at Google for 3 years and have 20%ed the entire time I've been there on: grpc-go, Drive, and Go tools (gopls, etc).
I think it's fantastic. The whole 120% thing is up to the individual: there have been times I've made it a 120%, and there are times when it's been just "take a friday off to work on other stuff". You end up getting less of your "job" done but my managers have always been supportive.
It's been great for sanity: some weeks/months feel just, like, meetings and chore work. It's great having that one day a week to work on a rockstar feature request in some fun project. It's also cool to work on your dream projects without the luck/physical move/whatever to get on the actual team. (you can effectively work on anything since no project is going to say no to free headcount)
It's also nice because it spreads your professional network in the directions you choose to spread it, rather than the more organic spread that your normal job entails (assuming luck and available are big drivers of where and which projects you "end up" working, rather than 100% your choice). So, maybe I don't work on project X today, but I can 20% on it and build up those connections, and later in my career I have a much better shot getting on the project. That agency is a nice feeling.
So, as far as the employee happiness goes, I think it's fantastic.
The myth that 20% time is dead was started by mchurch. He was at Google for six month and was let go. He then went on to bad mouth the company here and elsewhere online.
20% time at Google exists and managers are supposed to adjust the workload - it's not supposed to be 120% time. That said I think it would be hard (but not impossible) to a launch a 20% project to external users that doesn't have people working on it full-time.
I think of 20% time as error correction for management. Engineers can route 20% of their time to what they think is important, and not what management thinks.
Nobody from management told me to rewrite the primitive array handling in the proto libraries for Java. It just irritated me that they didn't do what I thought they should, so I spent some time fixing it.
Similar experience. I 20% on python...things. some python tooling, python3 was a major one, etc. I backed off recently just because most things I was involved in are wrapping up, but yeah, it was an active area for me. And also I am sort of 20%ing on my own project that is greenfield tooling with only dubious management support.
It's given me exposure to lots of stuff, I've learned more about a favorite language, met some cool people, etc. Not to mention earned a handful of bonuses directly related to my 20% work.
There's also now done tech debt reduction work that I've been tangential to that's clearly a value add, and it's primarily 20%ers.
Wondered if I could ask... is Google a relatively normal place to work? I currently do my own thing but I would love to work on projects that go far due to the amount of talent involved... but at the end of the day I'm just concerned with "office politics" I get the vibe that you have to "pick sides" which might just be an availability bias due to the news stories. I guess just relatively speaking, when talking with Amazon I get a different feeling. Recruiters probably aren't the best way to judge a company though, so figured I'd ask.
i guess I was asking this because I've never worked at Google and wondered about any downsides - e.g. distraction (the 20% project would seem to be the "cool" project, which makes the 80% boring) or politics (could be anything from jockeying for position on a hot 20% project or something else I can't even imagine because humans are petty)
I did a 20% project for Google Music back its early days, circa 2012. I was working on the Executive Compensation team at the time, but running a music blog called Indie Shuffle on the side. It's a long time ago, but I seem to recall naïvly hoping they'd embrace me with arms wide open and invite me to join their team + revamp their strategy.
I met with a couple members of their team who were open to entertaining me given my background with a music blog. I remember being really excited about it - and I spent a lot of time preparing a deck about how exposure provided by the Google Music blog could be used as leverage to give the platform legitimacy in the eyes of independent artists (something that SoundCloud was doing really well at the time).
A few senior leaders agreed to let me pitch my ideas, and after a fair bit of head-nodding, nothing actually happened.
I ended up leaving Google about 3 months later to take my music blog full-time (still up and running at https://www.indieshuffle.com, and eventually started a much-more successful music venture called SubmitHub - https://www.submithub.com). I count myself fortunate to say I have no regrets leaving Google.
Reflecting on the idea of 20% projects, I do appreciate that my managers gave me the opportunity to explore alternate opportunities within the company, and that the Google Music team was receptive to me poking my head into their affairs. I think it holds a lot of potential when it comes to retaining top talent that's at risk of jumping ship for something different, and made me feel like I was part of the larger company rather than simply stuck in a silo.
We're getting way off topic here, but I'm curious what your thoughts on the current state of the world on music blogging is. I was pretty big into following the music blog/hype machine scene in 2010-2014ish, and it seems like the other streaming services (particularly Spotify) have more or less killed that world with the exceptions of the biggest blogs, which seems pretty unfortunate to me.
Was there an exit interview? Was any one curious about your plans?
In my future perfect alternate reality:
Google Ventures, GoogleX, or even just thwarted & bored midlevel manager aspiring to become an angel investor, would offer seed capital to any one turning in their badge.
"Hey @jasongrishkoff, if you ever need some cash for an idea, please talk to me first."
In my mind, that parting offer is central to Silicon Valley's magic. There's an apocryphal tale about a much disliked boss offering seed money to mutinous employees. I thought it was Shockley and the Traitorous Eight, but I can't find a cite.
I just wanted to say thank you for creating Indie Shuffle - I discovered some of my favourite songs on there! I've found music discovery to be quite hard, and Indie Shuffle has been a great source for my taste.
Glad you plugged indieshuffle. I checked it out and love it so far. I'm going to make a point to listen to music through it for a bit and see what I find. So far all the songs I've heard are Soundcloud and it's hard to find any information on how/where you source the material. Is it soundcloud-only? Any interest or plans for integrating a site like Bandcamp that's more oriented around me eventually supporting the artist by buying something?
Great story, thanks for sharing. What is it like being on google’s exec compensation team? I had no idea that was a thing and would love to listen to any more fun stories (only if you’re allowed to share).
a) Learn deep learning on audio with a friend, via online courses, reading research papers, and re-implementing things. Then, to put the knowledge to work, I...
b) Joined a small bioacoustics project working with external researchers to level up their ML,
c) Developed some models and deliver some results to the external researchers, and, finally,
d) Got hired onto a new team doing ML on Audio full time, largely on the strength of recommendations from bioacoustics people.
IMO, 20% time is "just marketing" until you actually put in the personal effort to make something real out of it. Doing so is non-trivial, though.
There's a real risk of falling into a 'half-ass two things' pattern. It's difficult to do exactly one day a week on some project, then cleanly drop it until the following week. Context loss is a real problem. This year, I find myself looking out for 'low stress' times during my day job to do some deeper dive on bioacoustics and create a bunch of new stuff in a kind of sprint, rather than consistently setting aside 8 hours a week. It's hard to do a research 'sprint' in two areas simultaneously; it's better to let a research question take over my brain for a while.
(I also find that my personal limit for meaningfully tracking experiment outcomes is two model architectures. I tried three at some point this year and it was kind of a disaster.)
It's tough to motivate myself to do important-but-boring things like write unit tests in 20% time, which (combined with the context shifting problem) has often lead to pain down the line.
All told, it's a hard road, but very rewarding, IMO.
Ooh, I've been doing text-related ML for work, and have been tempted to do birdsong recognition for a hobby project over the break. Aside from that Kaggle competition, do you have any other resources you like?
At Google the joke was that 20% time was really 120% time since you still had to do your day job. I never personally saw a 20% time project turn into a product and I never felt comfortable taking 20% time. However, I did feel comfortable taking on what I saw as important work but which wasn't part of my job description --training myself up on it and becoming the go-to person for that work.
That part of the culture WAS consciously cultivated and I found it very valuable. It drew me up and made me more than I was. It was bound up with a culture of open sharing of ideas and cross-training. A notion that you could find compelling work at the company and shift your focus to that work/team.
There was also an effort to encourage new projects and ideas but they didn't go far enough. If I could give one piece of advice it's to create explicit approval and some serious financial incentives for people who start new products at your company. Treat projects like this in the way you'd treat an acquisition of the company making that product. E.G. if your company adopts a side project as a product, give the creators cash, respect, authority, and support to grow it into something great.
To be fair, 20% doesn't have to be a shipping product. I've had coworkers work on things like recruiting projects (running a puzzle/coding competition that ran on campuses around the US), employee resource groups, internal dashboards for various metrics relevant to employees, morale things like massive holiday lighting decorations around campus, and so on with 20% time. Those have all been very real things that contribute to company culture and morale and have "shipped" even if not to customers.
I worked at a different company that had implemented a similar policy (10% I think?), and generally the problem was the same. There was no process to formally accept projects, and no formalized way of having managers ignore the productivity loss when evaluating people.
I suspect the management liked the concept of having their smart engineers invent new products, but ultimately preferred buying companies. It somehow seemed less risky even though most of the acquisitions failed.
I'm not a Googler but wasn't Gmail a 20% project? (And maybe other Google products that I don't know?)
It seems to me that it would be entirely worth it to Google if 99.99% of 20% projects don't go anywhere, but every now and then, you get a big hit like Gmail whose newfound revenue completely eclipses the lost 20% experimental time from the 9999 other employees.
And even for those 9999 other employees, even if their 20% projects don't go anywhere, they are likely still hugely educational in ways that would make them more productive in the other 80%.
I'm not and never was a Googler but at Atlassian we the same policy of 20% time and I really do think it was super valuable. There are a certain number of developers who will muck around and waste their 20% time and there are certain (terrible) team leads and PMs who discourage the use of the time because they're anxious about their own projects.
But for the most part it contributed massively to the happiness of the developers. And the outcomes, in my opinion, were invaluable. It's not always visible from the outside, but Atlassian now has swathes of valuable internal tooling, built with love by developers who were invested in solving their own productivity problems.
The quintessential "20% project" is GMail but I think that misses the incrementalism that 20% projects really provide. Developers will absolutely take advantage of the time to improve their personal ergonomics, and everyone around them benefits from that.
> It's not always visible from the outside, but Atlassian now has swathes of valuable internal tooling, built with love by developers who were invested in solving their own productivity problems.
Can we get some of that same love put into the actual business products?
Another Atlassian here... I'll vouch for everything here, but I'll add on one thing.
> It's not always visible from the outside, but Atlassian now has swathes of valuable internal tooling, built with love by developers who were invested in solving their own productivity problems.
The flip side of this is that we have countless nonfunctional tools because the core maintainer has either left the company, switched to another project, or simply does not have the time to adequately adapt to new user/company requirements.
We do a great job encouraging innovation through 20% time, but we don't seem to be great at supporting these projects past the initial "hack it together" stage. The process can still be incredibly rewarding, but it is not without some shortcomings.
I wonder if any other companies implementing 20% time experience similar problems.
Honestly I wouldn’t agree to have 100% of my effort dictated by others. I never had a visa situation but even as a fresh out of college worker, I demanded a portion of autonomy. I wouldn’t be fulfilling what I could do for my employer otherwise. I get odd looks for some of the things I get enthusiastic but some of them work out. And I learn a lot more about things and in particularly the growth of new things from being fully responsible for some things. If you are a smart person who can code, you absolutely have the labor market power to take partial control of your time and effort. I would say you have an obligation as a thinking worker to cultivate yourself in this way, for the benefit of yourself, your coworkers, and your employer.
Xoogler here who left a few months ago. The policy totally still very much exists, although isn't "advertised" as much per say. Like, I think maybe 25% of googlers use 20% time and most are driving their own projects rather than working on some larger project they were pitched on.
One of the biggest positives of 20% time was just blocking off every Wednesday - no meetings, minimal distractions, etc - for my 20% work. It was pretty refreshing and kept me at the company for probably 8 months longer than I would have without the 20% time. The other great thing is that I was able to push a project I believed in but that everyone was afraid to fund. I had ownership of something I cared about. It was something that helped out dozens of teams, helped me shape the direction of the team, and heavily pad out my perf packet.
In general, folks are very protective of 20% time. My director didn't want to fund the additional maintenance cost of the project, but was respectful of the fact that it was 20% time. My manager and tech lead were in a similar boats.
The most visible impact was in the company's internal services. Many "scratch an itch" tools such as Google bus schedule lookup, employee tenure lookup, or massage-room booking originated in 20% projects. The "internal social media" of Google such as "moma badges" ("achievements" for e.g. writing 100 CLs or contributing to a particular fixit -- anyone could add new ones) and the internal joke image board "memegen" also originated as 20% projects.
This kind of modest, privste service was much lower-commitment than trying to single-handedly kick off a new external-facing service. Having all these volunteer-built projects around created a good vibe of being part of a community of engineers, each building whatever we needed to make our days better and sharing it with our coworkers.
I have no idea if it was good for their culture since I never worked there... but I reckon a lot of corporate software engineering jobs have the same luxury, informally. I work at a big US bank where you might not expect such freedom, but in reality there is great flexibility to do things outside of the normal day to day BAU, simply because it's not hard to deliver a sufficient amount of work in less time than you're expected to be in the office (or covid equivalent), and nobody keeps that close an eye on what you do anyway. Still, with all that freedom people don't always use it for good, they may actually just bunk off...
In the heyday of 20% time, Google had more money than they knew what to do with and the engineers were generally overqualified. Remember that in the early 2000s Google was famous for hiring people with PhDs and asking them build web servers.
Within that climate, 20% time was a great policy. Nobody needed "permission" to take a day out of the week to write a better source code management workflow or a config file VIM code completion for the internal codebase or improve build times or fix an annoying bug in another team's codebase, or even just generate good will for the company by contributing to open source projects. It was good for the company, and good for morale.
My experience while working there was that like most things the experience varied hugely depending on your manager and your current rating. For me, it was also kind of irrelevant since I'm one of those "weird" people that write code to relax in my "spare" time. Google didn't care what time of day you worked, I would find myself during the weekend trying out different things with map/reduce to see how different algorithms would need to be adapted and what not. Was that 20% time or just using the company's assets to indulge my curiosity? A bit of both I guess.
As a manager I'm pretty cognizant of the risk of burnout, and one cause of burnout is being hammered by a problem day in and day out. To avoid that I always coached my team members to have "three" jobs, one was their main job, one was their time to work on things that were important but not urgent, and the third was to indulge their curiosity and learn something new or further their understanding of something. But I also respected people who wanted to just "do their job" and go home, so for them we'd keep it simple about what was expected and how it was measured.
My working definition is something like “pre-product-market-fit” or “hockey sticking” growth. Once you saturate your market and have to fight for every percentage of growth (or are so big that new things can’t make a big impact even if they are exploding in their niche) then you’re no longer a startup
As someone who has worked for a long time at small startups (<30 people, <10eng) my "20%" projects (what other people are referring to "120%" here) that I've done on my own time have, in some cases, saved the companies I've worked for more than thousands of eng-hours and have taken less then a weekend. Attacking a core business problem from a new angle that no one else sees and designing a revolutionary approach to solving it that saves everyone time, money, and reduces stress is very common. Some examples of this from my previous job:
These were the ones that were easy to pluck from my job's monorepo and generally applicable.
As someone who is leaving a small startup for Google I can also add that it's one of the biggest compensation-y things in the offer. My manager seems supportive of 20%ing, the project I'm on is an offshoot of a 20% project, and the skip level I spoke to supported the idea of building tools.
So... From a management perspective the startups I've worked for have lost a senior engineer to another company more supportive of personal projects. This is very expensive as 1 senior engineer could represent 10%-40% of your eng team capacity at a small company. So, even just as an employee retention mechanism 20% of a salary spent on retaining your core engineering team and, possibly, obtaining huge benefits to overall company productivity is a worthy trade.
> even just as an employee retention mechanism 20% of a salary spent on retaining your core engineering team and, possibly, obtaining huge benefits to overall company productivity is a worthy trade.
thats a fantastic way to put it. as a businessman, 20% seems super high but who knows what the right number is. it's not 0%.
i'd be careful about IP in terms of "plucking from my job's monorepo" - 20% time work is still work that they paid for. but im sure you've considered that.
lastly - my qtn was not just about retention and saving money - it was also about culture - i do worry that politics and weird team dynamics arise when you basically play on a different team 20% of the time.
There never was any real 20% time. It was, as others are saying, 120% time, at least if you care about actually moving upward on the career ladder.
Paradoxically, this makes it a good policy _for Google_: if you're an idiot, you'll put in 120% effort and Google gets more work per dollar from you (and gets to deny promo because you "spent too much time on your 20% project", as well), if you're not an idiot, you will work 8 hours a day, but then spending 20% of time on things unrelated to your next promo completely dooms your chances of getting it. Some people used to take advantage of this, truth be told you're paid pretty well even at the senior level, and Google expects most people stay at that level indefinitely. Although then there's the question of why one wouldn't just take it easy at work and spend their spare time on hobbies instead that Google won't own the output of.
OTOH some people are taken in by the Google mythos, and they think spending 20% of their time on semi-random shit is _conducive_ to their career advancement. You get disabused of that notion after applying for a promo even once, which is when any experiments with 20% usually cease.
Exceptions are few and far between. To establish this for yourself, try to find if people around you promoted to Staff and above have any real 20% projects that are actually 20% projects, that is, that take substantial time. They don't.
I feel like at a place with expensive lab equipment, 20% time can go a lot further. Take smart people and give them a free day once a week to use all that expensive equipment they couldn't do on their own, and maybe magic happens.
I'll caveat my remarks first by admitting that I nearly completely ignorant of Google's 20% policy, but since it was widely advertised, my workplace made a similar statement that one half of each Friday should be dedicated to learning something new to benefit the team and thereby our careers. All I could ever think of after this proclamation was that I was to do as I was told 85% of the time and innovate 15%. As a developer, I offended that my innovations are devalued to a time slot. I was under the impression that I was expected to innovate continuously within the confines of the deadlines given. My two cents.
Googler here: I did a true 20% project once. I was filing a ton of bugs about Android internally and the process was crappy, so I found another person and we started working on an app to report bugs directly from the phone.
This went from nothing to a company-wide utility to a publicly launched app with full-time engineers (although I stopped my 20% contributions before that) [0]. So I think in this case we had a clear positive impact on the company that would not have happened without the 20% program. This app was launched by 2-3 20% engineers and 1 20% PM.
This was way outside my core job role (Developer Relations) but it was a great experience and I believe I was rewarded for it when it came time to apply for promotion.
People always seem to focus on the question of 20% time vs 120% time, but I don't think that question can be answered. Google does not have a time clock culture. I work about 45-50 hours a week. I have coworkers who work 30, I have others who work 70. When we had an office I got there at 7:45am, I sat near people who arrived at 11:00am. So basically time management is personal and so is how you choose to do a 20% project. In all cases the mandate is "get your work done".
I think it was great. The problem, in my experience, is that everything at Google is just too damn complicated these days. It's really hard to make an impact on a hard problem without an intense commitment. 20% context switching doesn't cut it.
Linters, code checkers, automation tools, IDE plugins, etc, have all been plucked. There's not a lot of low hanging fruit anymore.
My experience as a long-time Googler has been that people do 20% projects for two main reasons:
1) They want to change their role and view the 20% project as a means to an end. As an example, I worked with two 20%'ers that wanted to transfer from non-SWE to SWE roles. One of them has already succeeded, and another is hoping to be able to do so as soon as head count permits.
2) As a labor of love. As an example, we have an epitaphs page containing information about/farewells from people that have departed Google. It's not an officially supported project but is still quite popular. These kinds of projects don't necessarily lead to promotions (although the criteria for promotion have recently changed), but on balance they make people's lives better.
I used 20% time for things that were related to my work, but out of scope for my team. For example, I wrote a DNS library in Go for work and then I open sourced it [1] and used it to rewrite the standard library DNS client [2] as a 20% project. It actually worked out really well for me. The promo committee specifically called out my DNS project when they approved my promotion and ignored the stuff that I had been doing for my team.
[+] [-] deklerk|5 years ago|reply
I think it's fantastic. The whole 120% thing is up to the individual: there have been times I've made it a 120%, and there are times when it's been just "take a friday off to work on other stuff". You end up getting less of your "job" done but my managers have always been supportive.
It's been great for sanity: some weeks/months feel just, like, meetings and chore work. It's great having that one day a week to work on a rockstar feature request in some fun project. It's also cool to work on your dream projects without the luck/physical move/whatever to get on the actual team. (you can effectively work on anything since no project is going to say no to free headcount)
It's also nice because it spreads your professional network in the directions you choose to spread it, rather than the more organic spread that your normal job entails (assuming luck and available are big drivers of where and which projects you "end up" working, rather than 100% your choice). So, maybe I don't work on project X today, but I can 20% on it and build up those connections, and later in my career I have a much better shot getting on the project. That agency is a nice feeling.
So, as far as the employee happiness goes, I think it's fantastic.
[+] [-] ma2rten|5 years ago|reply
20% time at Google exists and managers are supposed to adjust the workload - it's not supposed to be 120% time. That said I think it would be hard (but not impossible) to a launch a 20% project to external users that doesn't have people working on it full-time.
[+] [-] rossjudson|5 years ago|reply
[+] [-] joshuamorton|5 years ago|reply
It's given me exposure to lots of stuff, I've learned more about a favorite language, met some cool people, etc. Not to mention earned a handful of bonuses directly related to my 20% work.
There's also now done tech debt reduction work that I've been tangential to that's clearly a value add, and it's primarily 20%ers.
[+] [-] robswc|5 years ago|reply
[+] [-] cbb330|5 years ago|reply
Naively, I imagine developers would get annoyed or bogged down having to skill up these contributors rather than viewing it as a “free hand”.
[+] [-] watermelon59|5 years ago|reply
My company tried doing the 20% thing but they put so much process around it that no one does it.
[+] [-] serial_dev|5 years ago|reply
Thank you for your take on these questions, your answer would be very much appreciated.
[+] [-] ffggvv|5 years ago|reply
so for a lot of people it doesnt exist.
[+] [-] swyx|5 years ago|reply
i guess I was asking this because I've never worked at Google and wondered about any downsides - e.g. distraction (the 20% project would seem to be the "cool" project, which makes the 80% boring) or politics (could be anything from jockeying for position on a hot 20% project or something else I can't even imagine because humans are petty)
[+] [-] alacombe|5 years ago|reply
All Google related projects.
[+] [-] jasongrishkoff|5 years ago|reply
I met with a couple members of their team who were open to entertaining me given my background with a music blog. I remember being really excited about it - and I spent a lot of time preparing a deck about how exposure provided by the Google Music blog could be used as leverage to give the platform legitimacy in the eyes of independent artists (something that SoundCloud was doing really well at the time).
A few senior leaders agreed to let me pitch my ideas, and after a fair bit of head-nodding, nothing actually happened.
I ended up leaving Google about 3 months later to take my music blog full-time (still up and running at https://www.indieshuffle.com, and eventually started a much-more successful music venture called SubmitHub - https://www.submithub.com). I count myself fortunate to say I have no regrets leaving Google.
Reflecting on the idea of 20% projects, I do appreciate that my managers gave me the opportunity to explore alternate opportunities within the company, and that the Google Music team was receptive to me poking my head into their affairs. I think it holds a lot of potential when it comes to retaining top talent that's at risk of jumping ship for something different, and made me feel like I was part of the larger company rather than simply stuck in a silo.
[+] [-] jcdavis|5 years ago|reply
We're getting way off topic here, but I'm curious what your thoughts on the current state of the world on music blogging is. I was pretty big into following the music blog/hype machine scene in 2010-2014ish, and it seems like the other streaming services (particularly Spotify) have more or less killed that world with the exceptions of the biggest blogs, which seems pretty unfortunate to me.
[+] [-] specialist|5 years ago|reply
In my future perfect alternate reality:
Google Ventures, GoogleX, or even just thwarted & bored midlevel manager aspiring to become an angel investor, would offer seed capital to any one turning in their badge.
"Hey @jasongrishkoff, if you ever need some cash for an idea, please talk to me first."
In my mind, that parting offer is central to Silicon Valley's magic. There's an apocryphal tale about a much disliked boss offering seed money to mutinous employees. I thought it was Shockley and the Traitorous Eight, but I can't find a cite.
[+] [-] jldugger|5 years ago|reply
What exactly does that work entail?
> A few senior leaders agreed to let me pitch my ideas, and after a fair bit of head-nodding, nothing actually happened.
I suspect successful 20% time projects are the ones where permission is not required, and don't need someone else to implement?
[+] [-] beaconstudios|5 years ago|reply
[+] [-] hamburglar|5 years ago|reply
[+] [-] ultimoo|5 years ago|reply
[+] [-] swyx|5 years ago|reply
so the tricky question is - why stop at 20% - why not 40% - or 80%? could you make a case for that?
[+] [-] donbrae|5 years ago|reply
[+] [-] sdenton4|5 years ago|reply
a) Learn deep learning on audio with a friend, via online courses, reading research papers, and re-implementing things. Then, to put the knowledge to work, I...
b) Joined a small bioacoustics project working with external researchers to level up their ML,
c) Developed some models and deliver some results to the external researchers, and, finally,
d) Got hired onto a new team doing ML on Audio full time, largely on the strength of recommendations from bioacoustics people.
e) I've kept hacking a bit on bioacoustics, including launching the birdsong id kaggle competition earlier this year. https://www.kaggle.com/c/birdsong-recognition
IMO, 20% time is "just marketing" until you actually put in the personal effort to make something real out of it. Doing so is non-trivial, though.
There's a real risk of falling into a 'half-ass two things' pattern. It's difficult to do exactly one day a week on some project, then cleanly drop it until the following week. Context loss is a real problem. This year, I find myself looking out for 'low stress' times during my day job to do some deeper dive on bioacoustics and create a bunch of new stuff in a kind of sprint, rather than consistently setting aside 8 hours a week. It's hard to do a research 'sprint' in two areas simultaneously; it's better to let a research question take over my brain for a while.
(I also find that my personal limit for meaningfully tracking experiment outcomes is two model architectures. I tried three at some point this year and it was kind of a disaster.)
It's tough to motivate myself to do important-but-boring things like write unit tests in 20% time, which (combined with the context shifting problem) has often lead to pain down the line.
All told, it's a hard road, but very rewarding, IMO.
[+] [-] wpietri|5 years ago|reply
[+] [-] rvna|5 years ago|reply
Also, what are you working on w/ Audio ML?
[+] [-] more_corn|5 years ago|reply
That part of the culture WAS consciously cultivated and I found it very valuable. It drew me up and made me more than I was. It was bound up with a culture of open sharing of ideas and cross-training. A notion that you could find compelling work at the company and shift your focus to that work/team.
There was also an effort to encourage new projects and ideas but they didn't go far enough. If I could give one piece of advice it's to create explicit approval and some serious financial incentives for people who start new products at your company. Treat projects like this in the way you'd treat an acquisition of the company making that product. E.G. if your company adopts a side project as a product, give the creators cash, respect, authority, and support to grow it into something great.
[+] [-] Arainach|5 years ago|reply
[+] [-] nitwit005|5 years ago|reply
I suspect the management liked the concept of having their smart engineers invent new products, but ultimately preferred buying companies. It somehow seemed less risky even though most of the acquisitions failed.
[+] [-] dheera|5 years ago|reply
It seems to me that it would be entirely worth it to Google if 99.99% of 20% projects don't go anywhere, but every now and then, you get a big hit like Gmail whose newfound revenue completely eclipses the lost 20% experimental time from the 9999 other employees.
And even for those 9999 other employees, even if their 20% projects don't go anywhere, they are likely still hugely educational in ways that would make them more productive in the other 80%.
[+] [-] mdoms|5 years ago|reply
But for the most part it contributed massively to the happiness of the developers. And the outcomes, in my opinion, were invaluable. It's not always visible from the outside, but Atlassian now has swathes of valuable internal tooling, built with love by developers who were invested in solving their own productivity problems.
The quintessential "20% project" is GMail but I think that misses the incrementalism that 20% projects really provide. Developers will absolutely take advantage of the time to improve their personal ergonomics, and everyone around them benefits from that.
But this is obviously very difficult to measure.
[+] [-] goalieca|5 years ago|reply
Can we get some of that same love put into the actual business products?
[+] [-] imtakmo|5 years ago|reply
> It's not always visible from the outside, but Atlassian now has swathes of valuable internal tooling, built with love by developers who were invested in solving their own productivity problems.
The flip side of this is that we have countless nonfunctional tools because the core maintainer has either left the company, switched to another project, or simply does not have the time to adequately adapt to new user/company requirements.
We do a great job encouraging innovation through 20% time, but we don't seem to be great at supporting these projects past the initial "hack it together" stage. The process can still be incredibly rewarding, but it is not without some shortcomings.
I wonder if any other companies implementing 20% time experience similar problems.
[+] [-] lanstin|5 years ago|reply
[+] [-] geekster777|5 years ago|reply
One of the biggest positives of 20% time was just blocking off every Wednesday - no meetings, minimal distractions, etc - for my 20% work. It was pretty refreshing and kept me at the company for probably 8 months longer than I would have without the 20% time. The other great thing is that I was able to push a project I believed in but that everyone was afraid to fund. I had ownership of something I cared about. It was something that helped out dozens of teams, helped me shape the direction of the team, and heavily pad out my perf packet.
In general, folks are very protective of 20% time. My director didn't want to fund the additional maintenance cost of the project, but was respectful of the fact that it was 20% time. My manager and tech lead were in a similar boats.
[+] [-] throwaway60011|5 years ago|reply
This kind of modest, privste service was much lower-commitment than trying to single-handedly kick off a new external-facing service. Having all these volunteer-built projects around created a good vibe of being part of a community of engineers, each building whatever we needed to make our days better and sharing it with our coworkers.
[+] [-] chamberecho|5 years ago|reply
[+] [-] dub|5 years ago|reply
Within that climate, 20% time was a great policy. Nobody needed "permission" to take a day out of the week to write a better source code management workflow or a config file VIM code completion for the internal codebase or improve build times or fix an annoying bug in another team's codebase, or even just generate good will for the company by contributing to open source projects. It was good for the company, and good for morale.
[+] [-] ChuckMcM|5 years ago|reply
As a manager I'm pretty cognizant of the risk of burnout, and one cause of burnout is being hammered by a problem day in and day out. To avoid that I always coached my team members to have "three" jobs, one was their main job, one was their time to work on things that were important but not urgent, and the third was to indulge their curiosity and learn something new or further their understanding of something. But I also respected people who wanted to just "do their job" and go home, so for them we'd keep it simple about what was expected and how it was measured.
[+] [-] Cederfjard|5 years ago|reply
Sorry for the tangent, but what’s the definition of ”startup” we’re working with here?
[+] [-] jmtulloss|5 years ago|reply
[+] [-] dnautics|5 years ago|reply
[+] [-] gravypod|5 years ago|reply
As someone who has worked for a long time at small startups (<30 people, <10eng) my "20%" projects (what other people are referring to "120%" here) that I've done on my own time have, in some cases, saved the companies I've worked for more than thousands of eng-hours and have taken less then a weekend. Attacking a core business problem from a new angle that no one else sees and designing a revolutionary approach to solving it that saves everyone time, money, and reduces stress is very common. Some examples of this from my previous job:
1. https://github.com/CaperAi/branchpoke
2. https://github.com/CaperAi/bazel_compose
3. https://github.com/CaperAi/pronto
These were the ones that were easy to pluck from my job's monorepo and generally applicable.
As someone who is leaving a small startup for Google I can also add that it's one of the biggest compensation-y things in the offer. My manager seems supportive of 20%ing, the project I'm on is an offshoot of a 20% project, and the skip level I spoke to supported the idea of building tools.
So... From a management perspective the startups I've worked for have lost a senior engineer to another company more supportive of personal projects. This is very expensive as 1 senior engineer could represent 10%-40% of your eng team capacity at a small company. So, even just as an employee retention mechanism 20% of a salary spent on retaining your core engineering team and, possibly, obtaining huge benefits to overall company productivity is a worthy trade.
[+] [-] swyx|5 years ago|reply
thats a fantastic way to put it. as a businessman, 20% seems super high but who knows what the right number is. it's not 0%.
i'd be careful about IP in terms of "plucking from my job's monorepo" - 20% time work is still work that they paid for. but im sure you've considered that.
lastly - my qtn was not just about retention and saving money - it was also about culture - i do worry that politics and weird team dynamics arise when you basically play on a different team 20% of the time.
[+] [-] m0zg|5 years ago|reply
Paradoxically, this makes it a good policy _for Google_: if you're an idiot, you'll put in 120% effort and Google gets more work per dollar from you (and gets to deny promo because you "spent too much time on your 20% project", as well), if you're not an idiot, you will work 8 hours a day, but then spending 20% of time on things unrelated to your next promo completely dooms your chances of getting it. Some people used to take advantage of this, truth be told you're paid pretty well even at the senior level, and Google expects most people stay at that level indefinitely. Although then there's the question of why one wouldn't just take it easy at work and spend their spare time on hobbies instead that Google won't own the output of.
OTOH some people are taken in by the Google mythos, and they think spending 20% of their time on semi-random shit is _conducive_ to their career advancement. You get disabused of that notion after applying for a promo even once, which is when any experiments with 20% usually cease.
Exceptions are few and far between. To establish this for yourself, try to find if people around you promoted to Staff and above have any real 20% projects that are actually 20% projects, that is, that take substantial time. They don't.
[+] [-] corobo|5 years ago|reply
Nobody just let you do passion projects! Get back to your work
[+] [-] mensetmanusman|5 years ago|reply
It’s the only reason we still exist; most of our sold products were side projects a generation ago.
[+] [-] jedberg|5 years ago|reply
[+] [-] dacracot|5 years ago|reply
[+] [-] habosa|5 years ago|reply
This went from nothing to a company-wide utility to a publicly launched app with full-time engineers (although I stopped my 20% contributions before that) [0]. So I think in this case we had a clear positive impact on the company that would not have happened without the 20% program. This app was launched by 2-3 20% engineers and 1 20% PM.
This was way outside my core job role (Developer Relations) but it was a great experience and I believe I was rewarded for it when it came time to apply for promotion.
People always seem to focus on the question of 20% time vs 120% time, but I don't think that question can be answered. Google does not have a time clock culture. I work about 45-50 hours a week. I have coworkers who work 30, I have others who work 70. When we had an office I got there at 7:45am, I sat near people who arrived at 11:00am. So basically time management is personal and so is how you choose to do a 20% project. In all cases the mandate is "get your work done".
0: https://9to5google.com/2019/03/12/android-q-beta-feedback-ap...
[+] [-] anonygler|5 years ago|reply
Linters, code checkers, automation tools, IDE plugins, etc, have all been plucked. There's not a lot of low hanging fruit anymore.
[+] [-] shmoo|5 years ago|reply
1) They want to change their role and view the 20% project as a means to an end. As an example, I worked with two 20%'ers that wanted to transfer from non-SWE to SWE roles. One of them has already succeeded, and another is hoping to be able to do so as soon as head count permits.
2) As a labor of love. As an example, we have an epitaphs page containing information about/farewells from people that have departed Google. It's not an officially supported project but is still quite popular. These kinds of projects don't necessarily lead to promotions (although the criteria for promotion have recently changed), but on balance they make people's lives better.
[+] [-] iangudger|5 years ago|reply
[1] https://golang.org/x/net/dns/dnsmessage
[2] https://golang.org/cl/37879