tstone.exit()

After more than 15 and a half years working for Google, I’m taking advantage of the Voluntary Exit Program (VEP). This feels a little awkward: I hadn’t been thinking about quitting prior to June 10th, but when I read Nick Fox’s email that morning I realized that now is probably the best opportunity to leave that I’ll get. I’ve been enjoying what I’ve been working on this year, and my team is at a good transition point, most of the way through the liminal space between Launched and Landed. It’s better that I leave now, while I’m happy and having fun, then to quit at some nebulous future time in anger and despair. “Don’t cry because it’s over; smile because it happened” as the proverb goes.

VEP is a decision-forcing function. Making a decision like quitting a job takes a lot of mental, emotional, and logistical effort. Breaking Up Is Hard To Do, and I’m not very good at leaving or ending things; I’m usually one of the last people at a party. Between working a full time job and pursuing other interests, “Take a hard think about whether I should keep working at Google” is rarely at the top of my priority list. But I’m the sort of person who values time off, and VEP looks a lot like a year’s worth of paid vacation, an offer I think should be taken seriously. While I didn’t particularly want to quit right now, I’m not certain that I won’t want to quit in the next year, and if I did quit in 2026 I would feel pretty foolish about working for ten months when I could’ve been paid almost as well to not work. If Google wants to incentivize cashing in institutional knowledge for a year of having fun, I’m happy to make the trade.

I’ll still be living in Boulder, so don’t hesitate to reach out if you’d like to spend time with me. I host roughly-monthly game days at my house (I’ll add you to my mailing list), and will now have ample time for bike rides, hikes, beers, and concerts. If you start feeling there’s insufficiently eclectic dress around the office, I promise to wear a weird hat if you invite me to lunch. And for all the Googlers who can’t hang out in person, point your favorite RSS reader at flwyd.dreamwidth.org. My journaling output dropped after I joined Google because there were so many corp communication opportunities; I’m hoping to be more active on my public blog now that I don’t have a default audience of Googlers scrolling through our internal water coolers.

I leave Google a little shy of 5800 days, which puts me in the 96th percentile among Googlers. There are 77 Boulder Googlers who started before me (19 were in BLD when I arrived), and I’m the 5th-most tenured Trevor. I’ve had close to 3,000 lunches at Google, probably 2,500 bowls of granola, maybe 100 massages, and 12 managers. I’ve visited 7 offices, occupied 16 desks, fixed around 1,000 bugs, acquired more than 900 Moma badges, and authored about 13,000 memes. I conducted over 300 interviews and more than 60 of them got hired, plus dozens more I said “yes” to in hiring committee (and a couple “no” decisions I feel very good about). I submitted more than 3,333 CLs and reviewed six thousand more; my deleted lines more than double my added lines (farewell Explorer client!). I didn’t stay on lit-hold, so I don’t know how many thousands of emails I received, though I leave with nearly 7,000 threads unread.

For everyone still at Google, be excellent to each other and Google on. I'll be posting a bunch of draft memes that I never found the right timing for. Upvote them, downvote them, or ignore them, I don't care. Just please don't assume they're a hot take on some current event; I'm just flushing my cache to persistent storage.

Shouts out

This is a woefully incomplete list of people and groups who’ve had a positive impact on me over the last 10000b years. Many of them are no longer at Google.

What’s next?

What’s immediately next is going to Burning Man. The VEP timing means my last day with full corp access is the Tuesday I was planning to hit the road, so that’s pretty convenient timing. I’ll then spend a month and a half in a liminal state with access to email, Workday, and benefits but no other corp systems. During that period I’ll pass the 16-year anniversary of seidl@ putting my résumé into the Google hiring system, marking 24 trips around the sun in my dance with Google.

I’m hoping to spend the fall relaxing and catching up on things that I’ve had trouble making time for with a full-time job: reading books, soaking in hot springs, ham radio in parks, playing with cats, learning cartographic tools, hacking on personal code projects, and installing clean energy systems in my house before IRA tax incentives are eliminated. I’m sure I’ll do some travel; maybe I’ll experience Carnival in Trinidad—I’ll have plenty of time to plan something once I don’t have a job :-) I’m hoping to keep funemployment going through summer 2026. My big worry is dealing with America’s ridiculous health care system where your employer will pay for your insurance if you work 90% time by taking every other Friday off, but you have to foot the whole bill if you work 90% time by working for 15 years and then taking 15 months off.

I’m not intending to retire permanently, though. Partly because I don’t think the earnings of 15 years at Google will cover my expenses, plus those of my parents, for the rest of my life and theirs. But mostly because there’s a lot of systemic work that needs to be done, and I hope that I can offer my skills to make it happen. When history books cover the 2020s, a big part of the story will be how humanity dealt with climate change, and I want to say that I made a significant difference in solving that problem. The current U.S. administration is cutting a lot of resources that people were counting on for climate solutions, but I think there’s a lot of good work that can still be done. And with this year’s big cuts to climate science, there may be a lot of smart and unemployed people in Boulder soon. I’ll be on the lookout for projects and companies this clever cohort of climate calculators concoct. What might a startup full of climate scientists create?

But why?

This isn’t the first time I’ve considered leaving Google. I almost quit in February 2015 when Google announced it would ban sexual content on Blogger. Not only was that an offense against free speech, it also shook my belief in Google as a reliable place to share content. I was working on Google Drive at the time; if Google can announce that years worth of carefully-curated naked pictures will be hidden in a month, why should users have faith in the longevity of files they uploaded to the product I worked on? Lots of passionate Googlers spoke up about that decision; Memegen was even disabled in protest. David Drummond and Rachel Whetstone held what I would describe as a tense town hall conversation where folks like Matt Cutts spoke eloquently about the issue. And remarkably, Google listened to us and changed its mind. I decided to stay because I want to work in a place where employees are passionate about defending users’ rights, despite social stigma and marginalization, and where leaders listen and course-correct when employees raise concerns. It’s hard to imagine that scenario playing out today.

The second time I almost quit Google was in March 2020. I was feeling pretty drained and disconnected in 2019—partly due to sleep challenges. I realized that most of my vacation since joining Google was oriented around events like Burning Man, and I hadn’t done any international travel in a decade. A trip to Iceland helped, as did a device to combat sleep apnea. But as the winter of 2020 wore on, I was still just generally exhausted. I felt like I’d crawled past the finish line of my 10 year Googleversary and I just wanted to take a couple months off, enjoy the spring and early summer, then find some way to volunteer for the election.

After five years at my first job, I’d quit and spent six months traveling and having fun before interviewing at Google, and the ten year mark here seemed like a good opportunity for what I like to think of as a “deep cleanse.” When an office email in mid-March told everyone to work from home for two weeks and grab whatever you need from your desk to be productive, I packed up my whole desk, figuring I’d wrap up a few things and give my notice before everyone came back. Well, 2020 turned out to be a lot more than two weeks away from the office, but my desire to quit also disappeared. First, a global pandemic lockdown is a lousy time to lose your health insurance in favor of traveling to exciting locations. Second, it seemed like the whole world had slowed down to my energy level: social norms about work hours, self care, and mental stress suddenly changed. Going for an hour-long walk around the neighborhood in the afternoon, taking meetings from your pajamas, and spending the first five minutes of a meeting checking in with each other were all suddenly okay in the workplace. Third, traveling around the country wasn’t a great plan in 2020, and if I was going to be spending several months hanging out in my living room, I might as well get paid for it.

The third time, I was really upset by the layoffs in January 2023. I considered resigning in protest. But the layoff severance package also seemed pretty sweet, so part of me wanted to stick around and hope to get laid off in the next round. 2025’s Voluntary Exit Program setup is basically an opportunity to raise your hand and say “Lay me off please!” I’ll be getting paid roughly a year’s salary to not work; as a time-off aficionado, that’s a tough offer to pass up. And unlike the folks who got cut in 2023 and 2024, I can say proper goodbyes, tie up project loose ends, and savor my last two months of free lunch, subsidized massages, silly team social chats, and delicious micro kitchen snacks. My team’s scheming out to use up our fun budget before I leave, too. And if it saves another K&I Googler from a layoff who goes on to help organize the world’s information then it will be good community contribution.

There’s more behind my decision than the desire to chill for a year, though. There are a lot of reasons, but I think most can be categorized into “Google’s not working the way I want” and “America’s not working the way I want.” The two are related.

Google’s not working the way I want

As I thought about enrolling in VEP, I reflected that I spend a lot more time talking about how Google’s not like it used to be than I do talking about the cool things Google’s doing. Partly this is because I’m entering my middle-age curmudgeon stage, but it’s also a reflection of the fact that the things I was personally excited about Google doing when I got hired aren’t the things that Google is especially excited about doing anymore.

I joined Google at the end of 2009 as a 30-year-old who’d fallen in love with the Internet in middle school in 1993. I was inspired by the values of the early ‘net: freedom, openness, inclusivity, education. A big part of the story of the Internet in the 1990s and early 2000s was about making information and communication freely available to anyone who could connect a computer to the network, and building infrastructure so that more and more people could do so. This was the environment in which Google was born and adopted the mission “Organize the world’s information and make it universally accessible.” It’s a mission I was excited by in my dorm room when I first visited google.com in December 1998, and a mission I still aspire to, even though you don’t hear it uttered as often these days.

The presentation at my Noogler TGIF was about Google Public DNS. The team explained that they’d identified that sub-par DNS servers run by ISPs were one reason that users had a poor experience on the web. Google’s DNS would make the web faster for users, and that was enough to make the project happen: Google realized that when the web got better, people used it more, would do more searches, and see more ads. That’s the kind of virtuous cycle that made me go home certain I’d joined the right company.

The environment at Google in the ‘00s is legendary, and I could feel that strong culture and values throughout the company. People were creative, irreverent, caring, quite generous, and committed to making the Internet and the world a better place. A big reason for the stability of this culture were company-wide meetings, which continually reinforce culture and norms. Larry and Sergey held TGIF every week, a meeting where teams could share something exciting they were working on (often long before it launched) and anyone in the company could ask a question to leaders, even if it was strongly critical of something Google did. It’s kind of wild that even as the company grew to 100,000 people, there were individual contributors that everyone felt like they knew—Waldemar, Blake, Jeffrey Oldham, and others—because they regularly stood up as a voice of conscience to a couple billionaires on stage. TGIF wasn’t the only meeting that provided this social glue, either. I really appreciated Eric’s quarterly talks where he walked through the slides shown to the Board of Directors and gave insight about the company’s strategic direction. And of course the Eng All Hands meetings where Amit would tell us how many Wikipedias worth of quality Google had gained over Bing that quarter, Nikesh would show numberless graphs of ads performance, and we’d be treated to a presentation about something ambitious like Google’s plans to build a bigger Internet to meet all of YouTube’s projected bandwidth needs.

One thing I tell interview candidates who ask if Google is a good fit for them is that different people thrive in organizations of different sizes. Some people do best in a startup environment: 10 people is great, 100 is getting a little strained, and 1000 is way too many: you can’t get to know everyone. Other people do great in a large company where there’s stability and the opportunity to move around, even if it means having to deal with more paperwork and process. I joined Google at about 30,000 people, but it felt a lot like a company of 1,000. This medium-company feel was in big part because of those all hands meetings, plus the culture of openness and collective dislike of unnecessary bureaucracy (pour one out for go/bureaucracy-busters). I also had the privilege of working in the Boulder office with about 150 people, and everyone got to know each other at lunch and other office events. But while the laws of organizational psychology are squishy, a community is bound to hit scaling challenges and Google’s no exception. There are too many people with reasons to leak for TGIF to share unlaunched products and candid thoughts from leaders. There are so many more things going on at the company that building the mental context for a meeting like Eng All Hands would be a challenge. At 100,000 people Google definitely didn’t feel like a medium-size company any more, and now that there are more extended workforce members than full-time employees it mostly feels like a big company that happens to have a quirky history than like a small town where everyone’s pastime is making the web better. There doesn’t seem to be a concerted effort to build medium-size-company culture and cohesion within product areas: the Knowledge & Information product area has close to as many people today as Google had in 2009, but I have far less of an idea what neat advances are happening in Search or Ads than I did a decade ago. Teams have evolved into their own ecosystems, which is good for focus but not great for Google as a connected community.

One symptom of Google being a decidedly large company now is the slow expansion of formal policies whose benefits are centralized and whose costs are externalized, leading to a steady stream of “ugh” changes to the leaf nodes in the org chart. To pick one example, two months after I arrived in the Boulder office as the weird hat guy, Helen asked if the Culture Club could use me as the star of “Can you beat Trevor’s hat day.” I set up a tripod and took some photos of me wearing silly hats (the art of the selfie was a lot more complicated back in the day), and soon there were fliers with my face on them on half the doors in the office, encouraging everyone to wear a fun hat on March 17th. It was a big hit, and a great example of office culture inclusion. In 2025, nobody’s allowed to post anything on the walls in any Google office, unless they’re required by law. Rumor has it this is an executive response to avoid unionization efforts, or maybe it’s just a global REWS team that found it too difficult to craft an approval guideline that xWF folks could implement. Either way, the weekly board game group can’t put up fliers encouraging people to come play Code Names with us. This is just one example of what feels like a growing trend of teams externalizing costs onto the Google community as a whole so they can optimize their own metrics. Everyone feels like their part of the system is working really well, but this builds wear and tear on all the other components and the system starts to run poorly.

America’s not working the way I want

… and that affects how I experience Google.

I was pretty unhappy about the U.S. election in 2016. This dissatisfaction wasn’t so much about the outcome, but about the way Americans treated each other. Folks had figured out that an effective way to get a lot of attention in the 2016 media environment was othering people. Donald Trump mastered this technique, stoking existing negative perceptions of all sorts of people. But Trump didn’t create this divisive environment in America; he found a leaking fire hydrant and figured out how to attach a hose. This culture of othering wasn’t limited to the right wing, either. Folks on the left were dismissing people’s experience out of hand based on social signifiers of class, education, and hobbies. People who’ve been trained on unconscious bias and cultural sensitivity would quickly jump to the conclusion that because someone supports Donald Trump they must also hold racist or misogynistic or xenophobic views. The Trump phenomenon didn’t appear from a vacuum. This cultural othering and antagonism had been on talk radio and cable TV for years. Trump’s ability to use Twitter and other social media to get everyone to pay attention to whatever outlandish thing he said brought this discourse to the mainstream. The tech world had spent the prior decade building tools so that everyone could add their voice to the conversation, so people spent the election deriding Americans they don’t know so they can be seen by their friends as loyal members of the tribe.

Google products played a role in this ecosystem of cultural antagonism. YouTube is one of the biggest social media sites, and a decade of development on engagement metrics and recommendation algorithms had the emergent property of optimizing for rabbit holes where someone without pre-existing negative thoughts about a topic could develop a strong hatred of some group or concept just by continually watching the next recommended video. None of the ML system’s inputs were “How’s this person’s mental health?” or “Do they better understand the world?” It’s like observing that people tend to stare intently at car crashes, then routing drivers through the most crash-prone streets with the expectation that accidents are what people really want to see when traveling. Google’s advertising products also played a big role: the most effective way to make money writing articles that are high on outrage and low on facts is to put it on a website with a lot of ads.

The internal Google community seemed to largely avoid this tribal antagonism up through 2016, in part because several factors tended to select for Googlers already aligned with one tribe. There was a significant amount of shock at Google after November 8th, 2016, and it felt like we were able to shelter and support each other in the aftermath. But as 2017 played out, there was significant social pressure to score points on the hot topic of the week. People started leaking TGIF so that a Google executive’s take on a Dory question about current events could bring 15 minutes of fame and clicks in the online media. Whatever short-term points the leaks won in some political or cultural dispute, the long-term costs were a Google where leaders and teams didn’t feel comfortable sharing thoughts and plans in confidence. The American social distrust of each other had reached inside Google and stopped us from trusting other Googlers.

Come 2024, the social game of tossing social media barbs at the other tribe wasn’t just on the external Internet; it was being felt on Memegen and other Google intranet spaces beyond “You were warned this is a contentious space” like rants@. In addition to a polarized American election, there are two major wars that are tied up in personal and national identity. Back in the 20th Century there was a view that it wasn’t polite to talk about politics or religion. But in the 21st Century, politics and religion have inserted themselves into people’s lives in such a way that not talking about politics means not talking about what’s going on in one’s life. For some folks, the answer to “How are you doing today” is going to look a lot like political discourse, because that person’s identity has gotten wrapped up in a hot button political issue. Google’s response to this was to discourage, chill, and often ban conversations about potentially-political issues in certain fora like Memegen. While from an institutional perspective this sounds attractive as a way to avoid people losing productivity over political arguments, it misses the human element of what employees are feeling: we’re distracted by political arguments already; making memes is a cathartic way of getting it out of our system so we can focus on work. The takedown process is also deeply unsatisfying. Whatever damage that was done to the person who reported the meme has already been done. The anonymous and indirect takedown process means the reporter and the creator don’t have an opportunity to have a conversation about the topic: where was the creator coming from, how did the reporter receive it, what’s the cultural context that each person brings to the story? This kind of context is how I help mediate conflicts at Burning Man, and it’s key to why restorative justice works. But instead we get a faceless process focused on minimizing risk and no real opportunity for learning and connection.

At the beginning of 2025, the Memegen takedown process, which purportedly intends to make the platform more welcome, made me feel decidedly unwelcome. I made a meme with a picture of a former president who’d just passed away and the text “Thanks Jimmy Carter for showing us all how to age gracefully and have a fulfilling retirement.” Aging gracefully and transitioning into different roles and life stages is something that’s important to me as a core part of a fulfilling life. As my beard turns white and my eyes strain more to read my smartphone in the morning, I’m conscious that I’m approaching Greygler status. It’s a stage of life when we can provide sage advice and wisdom from past experiences that rhyme with the present. In my lifetime, Jimmy Carter has been one of the clearest examples of the wise-old-sage archetype, and he won the Nobel Peace Prize for playing this role. I made the meme to honor a great man who set an example of living an active retirement, growing old with grace, and practicing influence without power. But in the political climate of December 2024, some unknown Googler assumed the meme was a dig at certain politicians who have not chosen to retire. ICMT therefore concluded that the meme was a political opinion, not allowed on the platform, and banned me from all Memegen access for three days. This escalation of conflict without mediation is exactly what Jimmy Carter worked to avoid, and as an aging Googler I felt excluded for talking about people who did a great job of growing old.

If Al Gore dies this year, it’s not clear to me that I wouldn’t get banned from Memegen for thanking him for raising concern for climate change. If the Dalai Lama passes on, would I get banned for thanking him for supporting Tibetans in a struggle for self-determination? If Aung San Suu Kyi dies in prison, would I get cancelled for celebrating her work in planting democracy in place of military control? If Jody Williams passes away, would I get in trouble for thanking her for making the echoes of war less brutal on people’s bodies? If a Nobel laureate passes away, I’m going to honor the work that they did, not just post a candle and their name. And if I were to get a lifetime Memegen ban for honoring three Nobel Peace Prize winners, I’m pretty sure I’d want to up and quit Google, since I wouldn’t have the intranet social space where I feel most at home. Since I can’t predict what kind of politicization anti-peace forces will do when these leaders pass on, I think I’d rather leave now while I can still post a farewell message on top of a funny picture.

Outside American political discourse, Google’s relationship with society and government has changed over the last decade and a half, and it’s not the teenage social rebel whose principled fights I celebrated a decade ago.

In my first eight years at Google, I can recall several prominent examples where Google pushed back on inappropriate government behavior or proposals. Two months after I started, Google announced the Project Aurora hacks perpetuated by China and that they would no longer provide a censored Chinese search experience. A few years later, Google helped rally the Internet to defeat SOPA and PIPA, flooding Congressional offices with calls and getting that threat to online expression stopped in its tracks. Google publicly pushed back on European “right to be forgotten laws” which enabled people to erase records of their poor behavior from the ‘net. Google signed the amici brief in the Supreme Court case that overturned bans on gay marriage, recognizing that winning equal rights for employees in their family life has immense business value. Google spent probably hundreds of fancy lawyer hours to fight a government surveillance request for Jacob Applebaum’s email and the secret gag order preventing disclosure of this overreach. And Google complained when Edward Snowden revealed the NSA was intercepting our traffic without a warrant, and acted quickly to encrypt everything to keep unwarranted spies out of our systems.

In the past eight years, I can’t think of many cases where Google has proudly and publicly pushed back on bad ideas from the government, aside from public statements about anti-trust cases and EU regulations focused on big tech companies. With smartphones in everyone’s pocket and social media displacing traditional journalism, it’s reasonable for legislation to step in to the online world in the 2020s where it stayed away in the 2000s when things were still developing. But it seems that Google’s approach is now a more traditional insider lobbying strategy, rather than helping lead the charge of public support. There are two ways to effect change in politics: mobilize money and mobilize people. My preference is for the latter, but Google’s shifted to the former.

One case where Google’s public silence on a topic struck me hard was the Trump administration’s executive orders regarding diversity, equity, and inclusion programs. There’s good research about the value diverse, inclusive, and psychologically safe teams bring to a business. For more than a decade we’ve recognized that some people arrive at Google needing more support because their background didn’t feature some of the privileges that I arrived at Google with. Google saw the value that affinity groups provided in supporting people who might otherwise not achieve their full potential, or even leave Google and tech entirely, so we funded and supported those groups. Part of Trump’s 2024 strategy was to run against DEI, presenting it as an unfair advantage given to minority groups, and preventing federal contractors from having DEI programs was one of his top priorities coming into office. But unlike Obergefell v. Hodges where Google and other tech companies presented the evidence of the business value in equity to the Supreme Court, Google quietly complied with Trump’s demands. women@ can no longer organize programs to mentor women in navigating a historically male field. VetNet can no longer mentor veterans in transitioning to the civilian world and private sector. If we can’t have formal efforts to ensure there are no missing stairs in Googlers’ career development, I’m willing to get off the stair I’m standing on so that I’m not crowding someone out who didn’t have my level of privilege in getting to that staircase in the first place.

An end and a beginning

Nothing lasts forever, and I feel I’m ending my Google career at a nice intersection on life’s road. In the month and a half since I decided to take the offer to leave I haven’t felt any regret, though there have been plenty of moments of tears and sadness, particularly when I shared my choice with my team. I’m going to miss a lot of people, a lot of traditions, and so many great tools that we’ve built. I’ll also meet a lot of wonderful people, start new traditions, and build some tools to solve the big challenges of the 2030s. I’ll still be around, don’t be afraid to come visit me on this new path.

Back to thoughts on Google
Back to my homepage

Created by Trevor Stone Last modified: August 07 2025 17:19:26
God requireth not a uniformity of religion. -- Roger Williams