Hacker News Reader

Comment by rgossiaux | Hacker News Reader

I don't quite understand. On the one hand you're saying that early employees should "demand market comp" and on the other you're saying that equity basically doesn't matter (and if you feel this way, it doesn't really make any sense to be joining a startup anyway). Are you conflating "compensation" with "salary"?

"Market comp" for a good engineer with several years of experience in the Bay Area is something like 250k - 350k at a Google or Facebook (with some variation in both directions here). It's not realistic, nor consistent with the market, for an early-stage startup to pay people this much in cash.

Early-stage companies should offer sufficient equity such that their employees should in expectation earn at least the same as they would at a public company. If their expected earnings are less than this, then more equity is absolutely a good solution.

View on HN

> Early-stage companies should offer sufficient equity such that their employees should in expectation earn at least the same as they would at a public company.

That's the joke! Nobody comes remotely close to offering enough equity that their total comp is equivalent!

Imagine a company that just raised a $1M seed round on convertibles at a $6M valuation cap. Now say they offer an "extremely generous" 2% equity package to their first employee. What is that 2% worth?

Well, the investors think that preferred shares in that quantity would be worth $120k. Of course, equity vests over four years. So your equity comp is... $30k/year. If it were preferred shares. But it's not, it's common shares. So it's worth even less.

You'd need to offer more like 10% to claim you are matching the RSUs people are getting at Google or Facebook. But at that point, you might as well call the person a co-founder. And maybe that's in fact the answer: add co-founders, not employees.

But this isn't what people are doing today. Instead they're convincing employees to take sub-market pay and sub-market equity to take a stressful job with almost no benefits.

Yes, I mostly agree.

There's one nuance that I've been thinking about lately that I haven't seen anyone ever point out before, which is that high volatility will make options worth more than they are on paper.

Here's a thought experiment: imagine that there are two employers on the market. One will always pay 200k/yr guaranteed and you can choose to work for them at any time for this wage. The other pays 100k/yr plus one Coconut per year, and Coconuts are currently worth 100k each. So far these are equivalent monetarily. But now let's add the stipulation that after one year, the price of coconuts has a 50% chance of going to 0 and a 50% chance of going to 200k each. Which would you choose?

The expected value of each of these job offers is still equivalent (after 4 years of working at each, you'll have 800k in expectation). But you should definitely take the second one. Why? Because after one year, if coconuts drop to 0, you can just quit and join the first company. Then you have a 50% chance of 100 + 200 + 200 + 200 (coconuts to 0) and a 50% chance of 300 + 300 + 300 + 300 (coconuts to 200) which comes out to 950k, which is better than sticking with either company alone.

This thought experiment fascinates me because it clearly shows the extra value that up-front stock grants have. The point is that you have some protection from downsides in the form of switching jobs, but no similar cap on the upsides, and the higher the volatility of the stock the more value this provides.

I'm not sure what the numbers look like when you view them through this lens, but it does mean that an equity grant of 2% of a 6M company should trade off for more than 30k/yr.

this also requires one exercises the cocounut option at the end of every year. If you don't and they drop to zero in year 4...

For the purposes of this thought experiment we're assuming that there's no price movement after year 1. In the real world this probability of failure in the future should be priced in to the year 2 price and doesn't affect the expected value anyway.

I think that's a good point, but it's just unlikely to actually happen with startup options because of the typical 90-day exercise clause and limited information.

The 90-day exercise clause and options instead of shares effectively forces you to buy the coconut, meaning that instead of a 50/50 chance of 0/200, if you plan on leaving before the event that resolves 0/200, then if your strike price on a coconut is $50, it's more like 50/50 chance of -50/250 because you expect to have to exercise when you leave. You can play with the exercise cost, but it does change your math quite a bit regardless. Now if startups gave actual stock instead of options or if they don't place a 90-day time limit on exercising vested options, this wouldn't be an issue. But almost all startups do it this way.

The second point is that the company often doesn't look dead until much later. You're just not going to be able to tell one year in, or two years in, or three years in. So you won't be in a good position to capitalize on volatility like you would on the public markets.

Then you have things like liquidation preferences that will skew your EV calculation, except they might happen after you join, but you may or may not ever find out that those gotchas are there.

As it stands, all of this just creates an incredibly inefficient market that requires employees to take badly informed, high risk bets, where they often don't have deep enough pockets to absorb the risk.

Both of these are good points. 90 day exercise windows are both unnecessary and terrible for employees and we should move to eliminate them as quickly as possible. I think the trend is slowly in that direction, but it really should just be a deal breaker for engineers.

The limited information is also a good point. Some level of saviness and understanding of your company and market can help with some parts of this but certainly not all.

If anything this may be a good argument for working at late stage private or smaller more volatile public companies. Then if you get lucky and get, say, Square from two years ago, whose stock has gone up 7x, you capture this upside (and if they do poorly, roll the dice again after a year, perhaps). This is quite mercenary and might not be the route that makes one happiest, though.

Yes!

Funnily enough, the fact that equity grants are "options" to purchase stock at a strike price less than the real share price is much less valuable than the optionality you describe of continuing working to earn the rest of a stock grant after more information is known.

Unfortunately I think it's very hard to pin a value on the optionality to continue working, and I haven't seen anyone mention it when considering joining a startup over a larger company.

But while FAANG equity has a 95% chance of still being worth hundreds per share in 5 years, the startup's equity has more like a 98% chance of being worth $0 in 5 years (based on startup survival rates). The expected value is essentially zero. That's why you can't make up for startup salary with startup equity.

When you offer 10% per employee, how many employees can you actually take on??

That was part of the comment: at that point it's more like a co-founder not an employee. So: very few obviously.

Perhaps more reasonable is 1-2% per for the first 5-10 hires, then 0.5-1% for the next batch, etc.

Also I really think these should be RSUs, not options. Employees are already invested in the startup by paying a premium in the form of reduced salary and opportunity cost.

If you give them Restricted Stock, though, they have to pay tax on whatever the 409(a) value is -- which is what the strike price would have been with options.

You can give them a signing bonus to cover the tax, but at this point it's easier and actually better for the employee to give them the cash separately and say "you can use it to exercise if you like".

Really the difference between options and Restricted Stock is measurable in dollars, so might as well just give people the dollars.

(Note that Restricted Stock is not quite the same thing as Restricted Stock Units (RSUs)... RSUs are for later stage companies. But the principles are similar.)

Sure; point being that employees ought not have to pay for their equity, because then it's not really compensation.

Yup, all the options for granting non-founder equity are absolutely awful from an employee POV.

Maybe YC could lobby for a sane way to grant equity to early employees without penalizing the employee?

I believe it was Sam Altman who had proposed a good solution to this, which is that you start off with a high number like that for employee #1 with decreasing numbers as the employee count goes up. If you are Instagram and become worth $1B with only 6 employees then everyone should benefit massively. On the other hand, if it takes 1000 employees to get to a great outcome, then obviously the founders and first employees will have been diluted by the addition of all the future employees required to achieve scale or an exit.

Yes, that’s a good way to go about it… Any educated opinions of what form it should take (resricted stock, rsu, options, etc.) to be most employee friendly?

250-350k is low for Google and FB. It's more like 400-500k if you are senior, 500-700k if you are staff. This is including RSUs

Indeed, and even well-funded startups can't afford to burn $400k on a single engineer.

They basically have to give out generous equity to compete, but they and the VCs would rather be greedy and dole out fractions of percents under the cynical misleading pitch that these scraps will be worth millions when the startup exits for billions.

To their credit, This scam did work for a while, shortly after a whole lot of early employees really did make millions on generous equity grants at early startups like Google.

Being an "early employee" means nothing now. You get the token 0.01% bottom-preference shares that will net you 0 in almost every imaginable scenario, and somehow this is supposed to cover the 200-300k/yr difference you'd get at a profitable established company.

How do you explain these numbers?

http://www.h1bdata.info/index.php?em=&job=Senior+Software+En...

Those are not Google and not Silicon Valley.

Try:

http://www.h1bdata.info/index.php?em=Google&job=Staff+Softwa...

And keep in mind that base salary (what's reported on that page) is usually about half of total compensation. (Bonus and stock being the other half.)

Also this appears to be the H1B database. I'm not sure how H1B salaries compare to the overall average.

> I'm not sure how H1B salaries compare to the overall average.

H1B salaries are typically lower than average. Why do you think companies spend millions lobbying for more H1Bs? To pay them more than average? :-)

Also, keep in mind Google only has to disclose base salaries for these H1Bs. For a staff engineer, most of the total comp would be in bonus pay and especially RSUs. They can easily be making $400k or more through those means.

> To pay them more than average?

No, to get a bigger talent pool.

Do you have evidence that people coming on an H1B visa get paid less than their non-H1B coworkers (at same level / same seniority / same office) at Google?

https://news.ycombinator.com/item?id=13579226 is one example. In addition to the article, HN is in near-unanimous agreement on this issue from the contribution of many H1B visa holders. I've seen it pop up many times.

I would suggest that while new hires of H1B might not get paid less, the market dynamics of not having as many alternatives would invariably lead to less valuable retention efforts by the employer. i.e. fewer raises, fewer promotions, smaller bonuses, etc.

As I understand it, H1Bs aren't too bad (transferability is a thing here), but other forms of visas are brutal in this regard.

H1b data doesn't include rsu or bonus. Many are getting 16-17% bonus and ~50-150k/year in rsu.

For a staff engineer, I'd say even $150k is very low. Even for an H1B, I'd guess $200k/year at least in additional pay beyond the base salary.

I work there and I'm surprised by the numbers you propose, they seem too high to me.

I'm not really interested in arguing about these numbers, it's not really relevant to my point. I was basing this off of https://www.teamblind.com/article/google-engineer---total-co... where 250-350 includes the majority of L4 and L5 engineers. Based on this your numbers look off by one level or so, at least for Google. But again, I don't think this is very important for my point.

rough numbers on http://levels.fyi

I always feel like every developer on HN is making like $200k-$400K without blinking, maybe that's true. Just for a perspective on these numbers, the top 1% of income in CA is $450K - https://www.reddit.com/r/dataisbeautiful/comments/8qhh3x/ann...

> Early-stage companies should offer sufficient equity such that their employees should in expectation earn at least the same as they would at a public company.

This would still be underpaying people for a few reasons: expectations are not risk-adjusted, it also doesn't take into account the timeline you get paid on, i.e. the ROI you would get investing your Google salary in an index fund and taxes make getting paid a regular amount over time more valuable than getting a large startup check.

Maybe you were taking these factors into account, but most startup equity offers are massive low balls.

> in expectation earn at least the same as they would at a public company

No, it should be multiples. Otherwise why is the employee taking on that risk?

I don't see why this follows. I said "in expectation". So some % of the time it's going to be 0 or thereabouts and some % of the time will be much higher. The best case scenario needs to be multiples in order to make the average work out, but I don't see any reason the average case needs to be.

Reply on news.ycombinator.com