Since returning to Meta in February, I’ve fielded a small deluge of calls from Meta people considering Google offers, Google people considering Meta offers, and non-aligned candidates considering both.

At face value, Meta and Google are similar companies. They both sell ads on the internet. They both build iconic consumer products used by a huge fraction of humans on Earth. They both encourage bottom-up innovation. They both built similar internal tools. They both use the same employee leveling system. 

However, product management at these two companies is quite different. There are certainly people who thrive in all environments, but I think that it is easier to be successful if you align your employer’s values with your own.

Quick background: Meta and Google are both gigantic companies and I obviously can’t speak for every team. I worked at Meta from 2018-2022 on speech, social audio, and language modeling, at Google from 2022-2024 on maps and language modeling, and now I am back at Meta working on the voice assistant for Ray Ban | Meta glasses. The scope of my current team at Meta is very similar to my last one at Google, so I feel like I have a reasonable apples-to-apples comparison of corporate culture in the applied AI space.

TL;DR

Meta and Google are both phenomenal technology companies where great PMs can thrive. If you are looking for convexity and growth at the expense of stress and pressure, Meta is probably a better fit. If you would like to prioritize work-life balance, stability, and job security, Google could be a great place for you.

Dynamism

Despite recently celebrating its 20th birthday, Meta is a dynamic company. Internally, someone put it really well: “in winner-take-all markets, if you can afford to buy call options, you do.” Whether it is social audio, short-form video, AI chatbots, or something else entirely, Meta leadership does not hesitate to bet the farm on new ideas, let them cook, and aggressively cut if they don’t pan out.

I experienced this myself with social audio. Clubhouse was blowing up, audio was the hot format, and I had the right experience. A couple hundred friends and I got invited to see if we could make social audio work in Facebook Blue. Less than a year later when it was clear we weren’t meeting expectations, our team was blown up. It was fun. It was exciting. But it was not stress-free. We woke up every day knowing that our careers depended on making numbers move in the right direction.

Google is different. Projects can span decades without being subject to evolutionary pressure. Reorgs seemed to be designed around fostering the long-term career growth of key people rather than aggressively pursuing a new trend in the market. When I pitched a new project to a Google VP, he responded, “This is great, but I want you to remember that our number-one goal is supporting our existing search business. I would rather do nothing than risk harming that.” This interaction encapsulates how Google thinks about change, which is likely correct from the perspective of Google shareholders but potentially not appealing to prospective product managers.

Note that there was one fairly public exception to this characterization, which I was part of. Sundar merged the DeepMind and Brain organizations by top-down fiat and mandated that a large fraction of their staff work on Gemini. I appreciated the conviction, but this mode of operating was unfamiliar to many Googlers.

Transparency

I think that transparency at both Google and Meta has declined, but Meta still feels like a flat, transparent organization. I regularly learn about other teams’ priorities through their Workplace posts and dashboards (and not infrequently contribute where I can) and have a good sense of what the company is up to. The expectation is that PMs lead frequent reviews up to the VP or C level and are held accountable for delivering on the conclusions. Managers are expected to communicate candidly with their reports about expectations for promos and ratings. Compensation is formulaic and predictable. Mark still holds Q&As where he takes reasonably candid questions off the cuff and directly engages with teams building across the company. The transparency is fun, but it comes at a cost: there is nowhere to hide. Your work will be visible to your leads, your peers, and your reports and it is difficult to take a back seat.

It is much harder to find out what is going on inside of Google. Unlike Meta, where almost all important company information is posted to open Workplace groups, Google operates mostly via emails and chats, which aren’t particularly discoverable. At the top, I never felt like Sundar was able to candidly answer TGIF questions the way Mark does at Q&A. And even within an org, it was never really clear to me how decisions were made or what work execs wanted out of their PM teams. In my 18 months at Google, I signed up for office hours with a VP a few times, but I rarely got any kind of leadership feedback on my work. Every time I brought up ratings or promos with a manager, I got brushed aside or asked, “why do you want to get promoted?” Compensation is subject to more manager discretion and can vary across a team controlling for level and rating. While this isn’t great for those looking to learn and grow, this organizational style makes it much easier to let work drift into the background when other life priorities need your attention.

Collegiality

The norms at Meta and Google are quite different when it comes to free expression. Meta feels like a quasi-academic, truth-seeking organization where decisions are made with data and dissent is encouraged. This environment can be quite unsettling to those used to a more consensus-based or non-confrontational culture.

Google culture is very reserved. Employees have often worked with each other for years and there is an expectation that impact will come over long periods of time if people avoid confrontation. Questioning priorities is generally not encouraged, which does tend to make for a more collegial work environment but can be frustrating for PMs who want to effect change.

Leadership

Both Meta and Google are bottoms-up companies. Ideas generally come from individual teams who build initial prototypes. If those prototypes get traction, others join the effort and entirely new products are born. Both Gmail (Google) and Marketplace (Meta) were famously side projects that grew into major components of the businesses.

However, it is very, very hard to self-organize large efforts. There is a famous internal doc at Google that compares the company to a slime mold for how it makes progress. Similar complaints fly around Meta with less pithy analogies. To extend the Google metaphor, almost inevitably, when two slime molds run into one another, there is conflict about direction. Even when one slime mold is chasing a single objective, sometimes it just can’t move fast enough to reach its target without a bit of extra nutrition.

Meta and Google adjudicate conflicts between slime molds very differently. Meta is Mark’s company. He has super voting rights and is a popular founder-CEO with the mandate to govern as he wishes. If Mark or his execs encounter two different visions for a product, they request reviews from the battling parties and make a call based on their judgment. This top-down control can cut both ways, depending on which side of the decision you’re on.

At Google, conflicts are not adjudicated. Teams can work on parallel projects for literally decades without any intervention from leadership, e.g. Google Maps and Waze, Brain and DeepMind, etc. This gives cohesive units, whether teams or entire divisions, the freedom to pursue their own priorities, but can also be frustrating for ambitious PMs who want to build products that require larger teams.

XFN

A common trope is that Meta is more PM-driven and Google is more SWE-driven. My experience at both companies seemed to confirm this.

At Meta, the expectation is that PMs are responsible for both ensuring the broader team is building something useful and that usefulness can be quantified and iterated on. 

At Google, projects were largely conceived and honed by engineers with PM serving as more of a supporting role. Both approaches have their merit, but I never could shake the feeling that Google could delete its entire PM function and not feel much in the way of repercussions.

Careers

Career progression at Meta feels much faster. The leadership ranks are stacked with young VPs who have done exceptional work over a relatively short period of time. Meta’s CFO, Susan Li, is representative. She was only 36 years old when she reached that position, which I suspect is one of the youngest in the industry. This attitude of rewarding excellent work independent of tenure trickles all the way down to L3 new grad engineers, who can achieve a promo per year for a few years straight if they perform.

Google is much more time-based. Each manager is given a promo quota based on the tenure of her team and promos (at least in my brief experience) tend to be given in a FIFO queue. Skipping that queue with exceptional performance does happen (there are a few young execs at Google that rose quickly), but it is much rarer. While Meta painfully cut 30% of the workforce several years ago, giving more room for promotions and new hires, Google still remains significantly overstaffed after the Covid hiring binge. As I was leaving, one VP told me that there was a 2-4 year backlog in L5-L8 promos that the company would need to work through before a normal pace could be reestablished. 

However, these career progression observations cut both ways. I have never been in this situation, but I believe that it is much easier to get fired for poor performance at Meta than Google, which should certainly be a factor for those for whom job security is paramount.

Comp

I intended to mostly communicate corporate culture in this post, but I do think that the different comp structures are worth mentioning.

Meta offers new PMs a salary, bonus target, and RSU grant that vests evenly over 4 years. Each year, a refresher worth approximately 25% of the value of the initial grant times the employee’s rating multiplier provided. These multipliers vary from 0x (“met some expectations”, which I think means you are getting fired) to 2.5x (“redefined expectations”, which means you did absolutely exceptional work). While the extreme ratings are rare, achieving a 1.25x or 1.65x multiplier is a common occurrence for employees who do good work. This means for the first four years, Meta employees make significantly more each year, independent of merit- or market-based salary adjustments and promotions.

Google offers the same except their RSU grant vesting is frontloaded, with 70% of stock awarded in the first two years (38/32/18/12). If a signing bonus is included, this makes Google first-year comp look really attractive and recruiters will communicate that refreshers will keep comp growing over time. Do not be deceived. Refreshers at Google are significantly smaller than Meta for the same level and role and performance multipliers are much scarcer. ~80% of Google employees receive “significant impact”, which offers no additional bonus or refresher. This means that the typical Google employee makes less each year, which is no fun at all.