I believe that most advice on choosing a startup to work for is wrong. Early employees at wildly successful startups suggest you assume the value of your equity is zero and instead optimize for how much you can learn. In this post I'll argue that evaluating how likely a startup is to succeed should actually be the most important factor in your decision to join one. As a former partner at Y Combinator, I know a lot about how investors do this. Now, as a founder and CEO of Triplebyte, I see how much less rigor the average job seeker applies to their decision and what they miss that investors would notice.
First you should be sure you really want to work at a startup. This is not the right choice for everyone. Paul Buchheit, an early engineer at Google, says,
“If you're happy working where you are, and you don't have any ambition to do anything else, you're probably going to get paid less and work more if you leave [for a startup]. If getting paid less and working more is unappealing to you, then I would recommend staying where you are!”
Joining a startup means giving up greater liquid compensation today for the chance of gaining other things that accumulate value over time. Things like personal networks, learning opportunities and equity.
You can learn new things and meet people working at any startup. However you will learn significantly more, build a stronger network, and accelerate your career trajectory much faster by joining a successful startup than an average one.
Successful startups attract the best resources in every category as they grow. They'll get the best investors, press coverage, and hire the best talent. The latter is especially important for your career. Initially, founders can only recruit talent from the limited pool of people who have the risk tolerance to join something early. As the startup succeeds, this pool grows and it becomes a magnet for the best talent. By joining early, you get to work alongside the smartest people and solve problems together without several layers of management between you.
It's also only by joining a successful startup early that you can get remarkably steep career trajectories, like Jeff Dean, Marissa Mayer or Chris Cox. Successful startups grow faster than they can hire experienced and qualified people to keep up with growth. This creates vacuums of responsibility, vacuums that existing employees who might be “under qualified” on paper can step in and fill. If you fill the vacuum well, you can keep repeating until you rise up into the highest levels of leadership of what will eventually become a public company. Even if you leave before reaching those levels, there's a halo effect around you now. Companies will roll out the red carpet to hire you, and investors will be keen to fund your startup ideas.
Once you've decided to join a startup, the first obvious thing to look for is any evidence that it is already succeeding. This is what investors do. Unfortunately, at the early stages there's usually not much data like this. When there is, one thing we learned at YC was not to be fooled by large absolute numbers. What matters most is the growth rate. A startup with $1 million in monthly revenue which has been flat for the past year is less exciting than one with $100,000 in monthly revenue that was at $0 six months ago. You care about the trajectory a startup is on, not just where it is at this moment in time. Joining Facebook in 2006 would have been a better choice than Myspace even though the latter had more users at the time.
Next, you need to evaluate the strength of the team and market. At the earliest stage, the team is just the founders. It's difficult to evaluate startup founders using traditional signals like college or work history. In fact, the kind of experience that looks good on a resume can be a negative signal for being suited to founding a company. Rising up the ranks at Google might mean you can only thrive in a well-structured environment working on a product with millions of users. Neither of those conditions apply to startups. However, it may also signal an outlying level of competence and competitiveness. Those are both good traits for a startup founder. That's what makes this so hard. There's not a single clearly defined shape of person you're trying to identify.
My advice would be to focus on how impressive the accomplishments of the founders are relative to their peer group and environment. Starting a company and convincing investors to invest is an accomplishment that contains signal. It contains more signal if achieved by a 19 year old college dropout from Milwaukee than a Stanford computer science graduate.
You also need to evaluate how well a founder understands the problem they're working on. Being an expert in the domain yourself makes this easy. You know exactly which questions to ask someone to call bullshit. If you're not, then take the approach of asking the founders to teach you about the problem they're working on. Be intellectually curious and see how well they can explain product choices they've made, how the market works, and who their competitors are. I particularly like asking what has surprised them since starting the company and which early assumptions turned out to be wrong. I would be suspicious if they claim every assumption turned out to be completely right.
Evaluating the relationship between founders is as important as evaluating the founders themselves. The top cause of startup death during a Y Combinator batch was cofounder disputes. We evaluated this very closely during interviews. We were looking for red flags such as talking over each other and disagreeing on important issues. It was a particularly bad sign if it wasn't clear who the CEO was. Likewise it was a bad sign if, when asked, one person meekly answered they were the CEO while the other glared at them. I'd recommend you request to meet the cofounders together to answer your questions so you can feel out this dynamic.
You can also get valuable data about the startup from the team that has been hired. They may know about any red flags or major issues the company is facing which the founders haven't told you about. This data is especially valuable because it is the hardest to get. Once you receive an offer, ask anyone you met during the interview to meet you for coffee. Then ask them hard questions such as, what are the biggest weaknesses of the founders? How have conflicts been handled? What major problems has the company faced thus far?
If you are an engineer, I'd also ask specific questions around how product ideas are generated and work is assigned. Do not look for a perfectly smooth and egalitarian process here. I believe early stage startups work better as benevolent dictatorships. You want at least one of the founders to have strong product opinions and be the final product decision maker. Early on, speed of action matters most to a startup, and decisiveness is a good trait in startup founders because it creates motion. Don't miss out on joining a great early stage startup because the founder seemed too opinionated and involved in product decisions early on.
You should also judge the quality of the product but don't always look for a perfect product. An imperfect product with bugs is fine if the team can explain why it still delivers a better experience than competitors. Moving quickly to launch software and sacrificing some amount of quality to gain users in a new or emerging market is a good strategy for a startup. This same tradeoff does not apply if the startup is in a crowded space with large incumbents who already have pretty decent products. Airtable is a good recent example of a product that has seen success launching in a crowded market by taking time to build and iterate on the product.
Next you need to evaluate how promising the market is. I won't cover how to do deep analysis of market size here. Estimating the true market size is an exercise composed of some parts rigorous analysis, guesswork, and persuasion. What is important is identifying the type of market the startup is in. There are two types of market for a startup. The first is a new market that is growing quickly e.g. blockchain. The second is an already known and large market with existing large incumbents e.g. real estate.
The type of market a startup is in should also impact how you evaluate the strength of the team and product. Being the first mover in a new and growing market has huge advantages which can cover up a lot of deficiencies in both the team and the product. If it is early enough, investors won't have funded many competitors yet. If you believe strongly in the growth potential of a market and find a startup that is currently in the lead you can afford to be more forgiving on the quality of the team and product. If the startup grows, both will be upgraded anyhow.
Operating in a market which is already known to be large and valuable will be less forgiving. There will be many well-funded competitors, and you'll need an exceptional team with access to capital and a correct vision for differentiation. Opendoor, for example, was not an average founding team. Co-founder Keith Rabois was an early team member at several startups that went public, such as LinkedIn, Paypal and Square. CEO Eric Wu had deep domain expertise after founding and selling a real estate startup.
Simply joining any startup won’t provide you with great learning opportunities or a strong personal network by default. A startup on a flat growth trajectory with a weak team offers neither. A consequence of this is being willing to move startups if the one you are currently at stops growing. I think startup employees should re-evaluate the growth trajectory of their startup every year. I’m conflicted in offering this advice. As the founder of a hiring marketplace, more people looking for jobs each year increases our market size. As the founder and CEO of a startup, I want my employees to stay with us for a long time and have faith during the inevitable rough patches.
There's a balance here. Every startup faces challenges, and if you're quick to jump ship you might give up a lot of upside. As a personal anecdote, my first startup was given an acquisition offer by Facebook in 2007. We declined in favor of taking another deal because Facebook seemed like a chaotic organization at the time with a lot of executive turnover and the conventional wisdom said it was overvalued. In hindsight, it may not have been a bad decision to join and stick through a temporarily turbulent time.
Predicting startup success is hard, and even professional investors fail at it most of the time. My advice here is intended to make you think more like them, in the hope that you'll build a portfolio of experiences over time that will maximize both your learning and your earnings. Good luck!
Thanks to Elad Gil, Ammon Bartram and Charlie Treichler for reading early drafts of this essay.