Based on my experience doing PM at small (<50 employees), medium (~300 employees), and large (50,000+ employees) tech cos, here’s what you can expect:

Disclosure: I prefer the PM role in startups, so I’m biased. But I’ve tried to be as objective as I could be.

Startup - Pros:

• Breadth. You’ll do a wider variety of things and get insight into more parts of running a business, which is very helpful if you want to start your own company later in your career. In a single day you might build a wireframe for a new feature, talk with a customer to get feedback on your design, help QA reproduce a bug, write marketing copy for the website, run some data analysis to understand how your latest new feature is being used, sit in on a partner meeting with business development, plan the next sprint, and do some research on your competitors. And that’s just before lunch!
• Impact (on the Company). In a small company, your work *matters*. If you do well, the company makes more money. If you do poorly, the company won’t succeed… until they fire you! You’ll get to know the CEO. You’ll get to know the investors. This again is helpful if you want to do another startup later. Same goes for company culture, the office environment, etc. You get a hand in building it, from the ground up.
• Freedom. As a general rule, smaller companies have fewer rules (HR, etc) and more leeway in how and when you do your work, what you wear, what you do with your workspace, how you behave, how you smell, etc. There are many, many exceptions here, but the stereotype is true: as a general rule you’ll have more leeway to do things differently if you’re not working for The Man.
• Small Chance at Vast Riches. I put this on the “pro” side but really this is a red herring. Very, very, very few non-founders, non-CxO people get huge exits from startups. If money in tech is what you want, then in dollars-per-year-worked there’s a clear advantage in working for a larger company. However, if you happen to be the first PM at YouTube or Instagram or… then you’re the exception that proves the rule. Most of the time, non-founder startup equity is worth anywhere from zero to a nice chunk of change (think: down payment on a house) but not life-changing money that you read about in the press.

Startup - Cons:

• Lack of Job Security. Most startups fail. If you have mortgages, kids, etc. this matters a lot. It’s also easier to get fired from a small company— if you don’t fit in with your CEO and the VPs, start polishing that resume!
• Amateurishness / Sloppiness / Ambiguity. Most startups have no idea what they’re doing. They learn over time, but in the meantime you’re unlikely to be exposed to best-of-breed practices in any area. Folks are just trying to figure it out. If you’re someone who likes the trains to run on time, then most startups will drive you nuts.
• You have to do everything yourself. Don’t like your desk? Here’s the address of IKEA… have fun! Don’t like the way the website looks? Here’s our Wordpress admin credentials… have fun! Don’t like the corporate logo? Feel free to use my Photoshop license… have fun! Got a question about your benefits? Here’s the 800-number for the insurance company… have fun! You get the idea— in a small company there’s literally no one to do the things, even the basic things, that you take for granted in a large company. Even for the basics like HR, janitorial, office supplies, etc. you'll often be on your own.
• Lack of PM Mentorship. If you’re joining a small startup to learn PM skills “at the feet of the PM masters” you are rolling the dice, not necessarily because startup PMs are bad but because (in my experience at least) startups don’t particularly put a premium on mentoring junior people. It’s up to you to figure stuff out and to get better on your own.
• No one has any idea what PMs do. Many people in many startups have never worked with a PM before. You’ll end up explaining your job role and its value, over and over again.
• Compensation. Startups tend to pay lower salaries and have stingier benefits, especially early-stage startups. Yes, you get a higher % of equity in a startup, but if you’re not a founder then it’s really hard to make more $$$ at a startup instead of a BigCo.

Big Co - Pros

• Impact (on the World). In a BigCo, you can PM an established product used by millions of people, without all the pain associated with building it up from nothing. You can just show up for work and make things better for millions of people. This is pretty intoxicating for a 20-something.
• Olympic-class nerd (and every other sort of) talent. At a startup, you’ll probably have a handful of superstars. At a BigCo, there will be hundreds of them, and you’ll meet them frequently and sometimes get to work closely with them. They may be hard to find among the hoi polloi, but if you’re good then you will find them and they will find you. Same is true, perhaps more so, for other disciplines— the best HR people I ever worked with were at a BigCo. The best lawyers. Amazing salespeople. And so on. This is great for building your network for your next gig.
• PM Support. At a BigCo there are usually many people who will offload tasks so you can move faster. A BigCo will have technical writers to author documentation, admins to help schedule large meetings, professional tradeshow people to design your booth, PR flacks to schmooze the press, etc. Not saying you always get the best of these disciplines, but if you just want to put your head down and write specs and work with developers and customers, this can be helpful.
• Name Recognition. People will know who you work for. This will often open business and personal doors for you. If you cold-call someone and say “I’m Joe from Google” or “I’m Mike from Microsoft” they will pick up the phone. Not so much if you’re Tim from some-startup-no-one-has-heard-of.
• Compensation. If you’re good enough to be promoted a few times, it’s a much more certain path to riches. VPs at most BigCos make way more $$$ for much less risk than the equivalent startup-focused career path. True, you also won’t earn Lamorghini money at a BigCo either… the top end is limited.
• Benefits. Although there are exceptions, most big tech cos have really amazing benefits and perks— and I’m using this broadly to include not just insurance and 401K, but also a great office where everything works, random discounts in all kinds of places, etc. Get into legal trouble on a business trip? Just call this number. Have a child who needs expensive autism treatment? Just call this number. Need a standing desk? Just call this number. Want a different chair? Just call this number. Want a cheaper rental car when you go on vacation? Just call this number.
• Company Stability, Breadth, and Growth. You’ll get job security but also the ability to plan for a longer-term career in one place. For example, you can work in one group for 2 years then transfer to another group without having to go through the hassle of switching jobs. BigCos also tend to have great management-training programs for good people, and clear career paths that you can grow into, if you’re good.

Big Co - Cons:

• Internally-Generated Large-Scale Randomization. Running a BigCo means placing a small number of nine-or-ten figure bets on core businesses, plus a bunch of smaller bets on smaller teams that might be material in the future. And inevitably those bets will change over time. Think about what happened when Google bought Motorola or Microsoft bought Nokia. Now think about all the smart people on those teams who saw years of their work go down the tubes because the acquirer decided to change strategy. Well, that kind of stuff happens *all the time* at a BigCo. Sigh.
• What you work on may not matter to the company. Unless you work on one of a small number of high-profile projects, your work won’t matter to the company’s bottom line. Even if you and your team does well, if another team is doing better or their VP has better rapport with the CEO, then you won’t get that extra headcount you need and your product might be cancelled. Several times in my BigCo career I needed to help the company shut down great products or features that weren’t material to the overall bottom line, even though they would have been successful (albeit small) standalone businesses. That was frustrating!
• Intra-Company Competition. Good management tries to prevent this, but in every BigCo there’s a zero-sum struggle for resources and executive attention. Too often this results in teams focusing on competitors in the next building instead of at different companies.
• Distance from the Customer. At a startup, if you don’t satisfy your customer you go out of business. Not so at a BigCo, where it’s easy to live in a bubble of other BigCo believers. It’s easy to lose touch with the actual users and buyers of your products, and easy to be insulated from the urgency of the marketplace. Lousy or unpopular products can get funded at a BigCo for a long time out of inertia, executive support, or many other reasons. (e.g. Google+, anyone?).
• Many BigCo PMs are Not Very Good. In BigCo’s there’s more of a need for inter-team coordination, stakeholder handholding, and project management— which can attract a particular kind of PM who tends to be patient, collaborative, but not particularly innovative nor feeling the urgency. I worked with many truly great PMs in my BigCo, but I also worked with many PMs who tried to make the transition from Project Manager to Product Manager but in the end didn’t add very much value. On their teams, developers and designers did what they wanted and PMs seemed superfluous. People who worked with those PMs came away thinking that all PMs are useless, which makes things hard for the rest of us. ;-(
• PM Culture May Not Fit You. In a small company, you’re likely the only PM or one of a small team, so you can shape the culture around your set of skills and gaps. In a BigCo, the PM culture is already established— you’ll need to adapt yourself to it… or leave. For example, you might like to build your own UI wireframes but if you find yourself in a company where that job is owned by UX designers (e.g. Apple), then you might be frustrated at the narrowness of your new position. Or you might not have strong technical skills, but be at a company (e.g. Google) where if a PM doesn’t have a strong tech background then they may struggle.
• Can’t Always Use the Best Tools for the Job. If your BigCo has a technology that does X and you’d rather use Y, tough luck. You’re using X. So you end up with Google not using Facebook’s open-source code, Microsoft not using Linux, and so on. By being part of a BigCo you’re signing up to use the company version of X, even if it’s not best-of-breed.
• Turf Battles. At a startup, there’s always so much more work to go around that people don’t usually have time for turf battles. In a BigCo, especially because the PM job by definition straddles boundaries between engineering, marketing, support, design, etc., you may be discouraged (implicitly or explicitly) going outside of the PM job description to get a job done. For example, in a BigCo you might end up waiting for a technical-writing team to finish the copy for your latest feature— but they won’t let you write it yourself, either! Ditto for marketing copy, sales training, UI wireframes, test plans, etc.
• Startup People May Think You Suck. Especially in the Bay Area where I live, there’s a snootiness towards people who work at big companies. It’s just not cool to work for The Man. For some people this is important. Other folks don’t care. But if you want to transition from a BigCo to work at a startup, be prepared for some anti-BigCo discrimination. I’m not saying this is always undeserved discrimination— there are lots of people at every BigCo who should never work at a startup— but there is a tendency among startup folks to assume that *all* BigCo employees are lacking in ambition and brainpower.

In summary, here’s my advice: try both a BigCo and a startup early in your career. Chances are you’ll feel more comfortable in one or the other. Follow that instinct. Don’t listen to the haters. They’re not you. Work where you like. Don’t be a hater!