Naval Ravikant, AngelList: eventually startups will be everywhere
AngelList is the largest startup community on the planet, connecting tens of thousands of startups with investors, media, potential hires and more. AngelList Talent, job marketplace, place up to 100 candidates into new jobs every week – a level I can only hope to reach one day at djinni.co.
AngelList and its co-founder Naval Ravikant have been inspiration for me quite a while. While in SF I couldn’t miss a chance to speak with Naval in person. Here’s slightly edited version of the interview, in Naval’s own words.
On coding culture at AngelList
Almost everybody in the company has technical background. Even if they don’t code today or code anymore they have to understand code, they have to respect code and automation.
This is important because non-technical people have a very hard telling what’s possible and what’s not and we’re in the business of technology so it’s important to understand capabilities, outcomes, tradeoffs, all those kind of things.
If you don’t have technical background you’re not going to have a credibility with the developers when you talk to them and when you work with them. We just think it’s important that everybody have some technical training and background. Even though not everybody is coding but even our deals team and our designers all have GitHub accounts and they go into the code base.
On “one-man startups” hiring philosophy
We hire people and expect them to be very self-sufficient, we call it one-man startups.
We want them to be able to do some design development, some product management and do their own customer service for a long time. They largely choose what to work on out of a set of central priorities and we don’t track when they’re in the office. It’s all tracking outputs not inputs.
Many people can’t function in that environment. They want to be scheduled. They want specs written down for them. They want to know when to show up, they want to be tracked, they want active feedback. Those people don’t do well at AngelList so we end up letting them go.
On product-market fit for AngelList
Right now it is kind of a dangerous phase for us. We probably hired more people in the last few months than we’ve ever hired in that short of a time period. We’ll see how it goes.
We have product market fit in our talent market which is placing a hundred hires a week. Then we have this great online investment product with syndicates and startup fundraising that people are using and it’s growing.
I think we have product market fit but there are definitely things we could do better. We still want to stay small. We still want to stay in the 20, 25 person range while we nail it.
I don’t like having managers. I’ve had companies before with more people [than AngelList] and with explicit management and there’s a couple of things I’ve discovered.
Number one, I hate management. I don’t like doing it. It’s stupid and boring.
Number two, I think the best developers don’t like being managed. They don’t like being told what to do. They want to have their creative freedom because creating products is a creative exercise.
Number three, the moment you’re working through other people just your efficiency slows down a lot. You spend a lot of time in meetings; you have a lot of mistranslations.
One philosophy I have at AngelList is I would rather have someone working on the wrong thing and do it their way and be motivated, than working on the right thing and do it my way and not be motivated. The reason is because I think that management’s ability to predict what to work on or what’s going to be useful is actually very low.
I don’t think I’m any smarter than any of my developers.
If I don’t have that ego about it then they’re just as likely to be correct as I am. The only thing that’s important for me is to communicate to them what the company priorities are, what the vision is and to give them all the data so that they make the right decisions. They’re just as smart as I am.
I would rather they make the decision, they’re bought into it and then they code it up in the best way possible as opposed to lots and lots of meetings where we argue or people tell them what to do from above and they sort of do it reluctantly.
Knowledge work is the ultimate fuzzy work. A good developer can be ten or a hundred times as productive as a mediocre developer and it’s just a question of how into it they are. Do they want to solve the problem or not?
On why startups are more productive than bigger companies
There’s a notion in economics called the principal agent problem where the principal is the owner and the agent is the person who’s actually doing the work and getting a cut. The principal agent problem basically says anytime you have an agent the agent optimizes for themselves rather than for the principles and interest. You just end up with all kinds of perverse behaviors.
Classic example is government.
The reason why startups are so successful compared to big companies and they’re so much more productive is because you have a small group of people who consider themselves to be principles. They are very few agents. The more managers you add, the more agents you are adding into the startup and so I think the less efficient you’re going to be.
On scaling an organisation
Let’s assume I have 10 developers and each one takes 5% of my time in management, right? That means I’ve only got 50% time left to do anything else but if I’ve got 5 developers who are equally good as these 10 but they take 0% of my time management or 1% then I have so much more time left. My organization can scale so much better.
I think there are a lot of companies that scale up with the number of employees.
The one common trait that they have is that they already have product market fit. Their product is basically working. The market wants it the way it is. Now their job is to scale it up, get it to more platforms and so one. You can’t screw it up at that point.
If you’re still trying to do something, innovative things like you’re still trying to say, “Well, I think the product’s mostly there but not quite,” or, “I can’t yet get people to adopt this in mass or I haven’t figure out how to make money out of it.” Those kinds of things have to be done by a small number of principles. If you get too many agents involved you’re not going to solve those problems.
I think premature scaling really hurt startups that way. They hire too many people before they figure out how to solve the problem.
Startups don’t pay well and they’re incredibly brutally stressful and difficult. Only do it if you really have a great concept that you really want to work on, product that you’re excited to build and one where you think you either care than anybody else or because you have some deep specific knowledge about it that other people don’t.
Startups are very competitive. Anything you build you’re going to have ten competitors or more and so you have to be the best in the world at it and care more than anybody else if you’re going to have a chance in success.
I wouldn’t do a startup just to do a startup. If you really care about something and you want to see it exist and you think you know it well and you’ll stay up nights when everybody is sleeping where you’re thinking about it or working on it then go do a startup.
There’s no single formula.
For every piece of advice I could give you there’s a counter example or an exemption and this is probably a big one. That’s why I kind of say at the end of the day just only pick something you really care about enough. If you care about it you’ll become world class expert at it.
I don’t think there’s any single right formula to prevent yourself from wasting time and I know that all of these methodologies about being lean and testing and all that stuff, and you can do all that.
Think of it this way, if you just came up with the idea in 3 months or a month then you’ll probably get bored of it in a month or 3 months. If you start it with someone that you only met last week you’ll probably start fighting with them after a week. The longevity really matters.
The longer you’ve been thinking about the idea, the longer you’ve been working in the space, the longer you’ve been working with the person that you co-found the company with, the better and more likely you are to be able to predict how it was going to be going forward.
That said you don’t want to sit on your hands forever 10 years on the job because it would just get spoiled or really you’re bored or lazy or whatever. It’s a fine balance.
You try every trick in the book at the end of the day there’s a ton of luck involved.
A lot of doing startups is just someone who really happens to know space extremely well and is really passionate about it just coincidentally starts working in the space just as the space takes off and becomes 10x bigger.
It’s much easier to avoid mistakes than just to do things right.
Don’t do something just because you think it will make you money but you don’t really care about it. Don’t do something because everybody else tells you it’s a good idea. You have to find if you think it’s a good idea.
Don’t start a company just because you feel like you have to be an entrepreneur. Don’t start a company with someone you just met.
Don’t start a company in a space where someone’s already won and then you think you can do it just a little bit better, right? Like, “Oh, that’s Facebook. They’ve already won but mine’s going to be better on mobile,” right? Don’t do those kinds of things.
You have to take big risks, almost like what I call 10x risks or Peter Thiel calls a “secret” – something you believe deeply that other people don’t. You have to be right and everybody else is wrong. Those kinds of things only come from deep knowledge and passion.
On creating a startup vs joining one
If you haven’t found that but you still want to be involved in a startup scene go join somebody else’s startup. But it’s better to join someone who’s already found a product market fit. They’ve already created something the market obviously wants.
We don’t do Agile, we don’t do Scrum, we don’t do pair programming, we don’t do Lean. We don’t have any religion.
Here [in Silicon Valley] good developers want to work for stock. They want to have a piece of the upside.
At my last company vast.com we have like a hundred people in Serbia and a lot of them were really good. I remember in Serbia when we gave our people their stock they thought it was a Ponzi scheme because we said, “Oh, these are stock options. You have to pay a little bit of money to buy a stock which you do here for tax reasons.” Over there it turns out people just think, “Oh, you’re taking my money. You said you’re giving me something but you’re asking for money from me.”
I understand where the suspicion comes from and yeah, in Ukraine it probably makes more sense for people to be paid in cash.
We actually have a graph on our site at angel.co/salaries with all salaries and jobs by salary. It’s pretty high.
If you’re a good developer you’re making in the $80K to $120K range. A lot of companies now pay up to $140-150K a year. And that’s because the big ones like Google and Yahoo are paying $250,000 a year or more, when you take the full package into account.
The supply of good developers is pretty fixed and the demand is just going crazy. We are seeing a lot of developers come out of these academies like Hackbright and HackReactor and DevBootcamp and so on.
Those are performing great service to society because they’re training more and more people to be comfortable with computers, but a 12-week training course is not going to turn you from a non-developer into a great developer. That’s going to take years and years of coding but it’s a really good start.
Unfortunately the retraining is still only being done kind of people mostly their early 20’s or late 20’s. No one who is like 55 and has been laid off from a manufacturing job is going to learn how to code at Hack Reactor but it would be nice. It would be good if more of society saw that as a viable path.
If Kiev programmers could come to the U.S. or go to New York or go to London and get established a little bit, their salary would be five or ten times or twenty times what it is in Kiev.
Now unfortunate cost of living would also be five or ten times higher but overall they’d have a much better quality of life. One of our best engineers in AngelList and one of the two people who originally built most of the site is a Serbian guy.
10 years ago you couldn’t have even made a living as a developer in Kiev unless you were in a very small set of people. Today you could go to Berlin; you could break out of a larger startup scene.
I think being mobile is a big deal. The best thing that ever happened to me of course is that my parents came to the United States when I was kid and then the second best thing was I moved out to Silicon Valley from New York.
Being in the right location matters a lot. Do whatever it takes to get exposed to more of the world. If you can even make a trip to Silicon Valley for 3 months do it. If you can go to New York for 3 months do it. Don’t just stay in your hometown.
It’s very unlikely that you were born in the best place in the world to do what you’re doing.
If you move around a little bit, try out a bunch of different things, try Berlin, try London, try New York, try San Francisco you may not be able to stay in a given place because of visa issues but just the exposure that will give you, the context, it will give you the mindset it will give you will be invaluable.
I think eventually startups will be everywhere. There will be startups in every city and that will really happen when we’re more cloud connected.
