Has anyone stopped to think what a massive failing of the startup part of the industry this is? Practically everything I read online indicates that if you consider your stock options to have any value at all even in a moderately successful company, you are a major sucker and about to get exploited.
Surely this must reduce the quality of the talent pool available to new startups, as the experienced developers conclude that other options are a better use of their time.
First of all, if you go looking for market inefficiencies in tech hiring (across the board, not just at startups), you will find lots of them. Software development hiring is folkloric; traditions handed down from Sr. Mgr Software Developer to Associate Developer tracing back to the beginning of time (1982 or so).
Second, regarding the talent pool available to employers, two factors confound the analysis: the first and by far the strongest is stated preference vs. revealed preference --- to wit, good developers will make large concessions on comp in exchange for working at companies that seem more fun; the second is that software developers are as a demographic cohort terrible at negotiating.
> software developers are as a demographic cohort terrible at negotiating.
Yep. It's no real surprise that coders are mostly men with poor social skills, while HR is mostly women with good ones, most of whom those men find attractive. Classic Valley symbiosis.
HR people do not as a rule do salary negotiation. You have to be a particularly "special" degree of bad at negotiating to end up out-negotiated by an HR person.
I am sure there are companies that, by outward appearance, do have candidates negotiating with HR people after the interview is over. Step 1 in handling negotiation with those companies: realize that you are not negotiating with HR.
This is a weird bit of advice. From my experience I have to assume you're saying "HR isn't the decision maker when hiring in elite tech companies" but the fact of the matter is HR/Recruiting is going to present the offer to most people, and it takes a career worth of preparation to move the conversation beyond that offer.
Well, I'm happy to have moved you a "career's worth" of wisdom forward in a single comment. You aren't negotiating with HR. HR does not know what you do, and HR's best idea of what you're worth comes from those ridiculous salary survey sites.
This is a cogent point. HR may be your point of contact for your salary negotiation, but they are not the decision maker or barrier. Ask for more money and they will ask someone else for more money on your behalf. HR is almost never the enemy in less-than-huge companies (and even then only moderately at worst).
I've worked for 2 large technology companies. The first was a big one down in Southern California and HR there was as you describe.
The other was a big one in San Francisco, and their HR was insanely powerful... for some reason. It was quite a shock to me but to a lot of others used to Bay Area startups they made it seem like the norm.
So I guess my point is that not all HR is alike and there is probably some truth to this HR negotiating business.
But they are the ones having the actual conversation and working the rhetoric to close a deal. They aren't Deciders, but they are Negotiators ("salespeople").
Most of the software developers I've worked with in my career have had very good social skills, those that didn't, were poor developers as well... So are they bad negotiators because of they lack negotiating ability, or actual ability?
Is there data on the social skills of IT, HR, Gender breakdown etc... Perhaps what you are refering to is a U.S (?) phenomenon?
The correct answer is that developers do not have, and are resistant to, unionization. As individuals they'll always have very little bargaining power, and most developers are pretty easily replaceable (especially before they're hired).
In other words, you're telling me both people and organisations are imperfect?! Shocking.
Sarcasm aside: really, it is shocking how SW engineers could come to think that any of the inexact, data-free, human-judgement-driven sides of business are optimal--simply from the theoretical argument that a market is involved. A little economics is a dangerous thing.
I suggest starting from the assumption that everything can be improved, unless proven otherwise.
> Has anyone stopped to think what a massive failing of the startup part of the industry this is?
The failing is not that the employee equity math rarely works out, it's that the "industry" is so focused on equity. It often falls short as a recruiting tool (a significant number of prospective employees are clued in to the fact that it's likely to be worthless) and it's usually a poor retention tool as well (just look at startup turnover and the number of employees who don't stay with one company long enough to fully vest).
The startup value proposition today is actually quite compelling in some cases. Employees, many of them young and without significant real-world experience, can earn six-figure salaries working at companies that, without outside investment, could not sustain themselves.
Too much capital chasing too few opportunities has given many startup founders the ability to raise capital on terms that are insane. I mean, you have entrepreneurs raising million-plus convertible note seed rounds with caps that make absolutely no sense. Where does all that cheap money go? For many if not most startups, one word: salaries.
If you're being paid $120,000/year plus benefits to work on a CRUD Rails app at a startup that probably won't be around in five years, you should forget about equity. You have already won the lottery.
How is $120k/year to live in a top 5 most expensive city in the world to work harder than 95% of the people on the planet on a boring CRUD app winning the lottery in any stretch of the imagination? That sounds terrible.
Try earning $25,000/year working two jobs in a top 5 most expensive city in the world.
Working in tech may be unjustifiably glamorized, but you have to be incredibly out of touch with reality not to realize that earning six-figures plus benefits working 8-10 hours a day in air-conditioned buildings for employers, many of whom will feed and transport you, is something that millions of Americans would give everything for the opportunity to do.
Precisely. Work in a boiler room, or hike sacks of grain 20 miles on your back, or work in low end food service, or go scrub toilets 60 hours per week, then come back and say software developers are working harder then 95% of other workers.
80+ hours per week? Are you joking? Most startups aren't 9 to 5, but please name a single startup where you believe employees regularly put in 80 hour-plus weeks.
Even in professions like law and investment banking, where employees do have to work grueling hours on a regular basis, the "80 hour work week" is largely a myth. I think medical residents are one of the few groups that really puts in these types of hours consistently.
Yes. And from what I've seen, the companies where developers regularly do put in stupid hours aren't doing it because it's effective. It's a sign of dysfunction. E.g., a competition to be seen as the toughest, or a manager who can't really evaluate productivity other than by counting butts in seats.
hah, don't get me wrong, I highly doubt anyone really "works" a 80 hour week. I personally work 35 given that I take an hour and half for the gym + lunch everyday. And out of that 35, I probably spend 10 reading HN, learning new tech, doing personal emails, and other not exactly job-related activities, so 25 I'd say total of real 'work'.
so I was probably exaggerating a bit, but I stand by my original comment that many of these overworked, underpaid startup employees lead absolutely miserable lives and work harder than 95% of people on the planet.
For reference, the US in general works more hours per week than any other industrial nation. Hunter gatherers worked only 15 hours/week. Most impoverished nations work very few hours per week. The only people who beat them out are sweat shops in Southeast asia
"Employees, many of them young and without significant real-world experience...."
Getting that first job on your resume can be quite a trick.
Just before he emphasizes, with italics, that this is only "compelling in some cases".
Because of the circumstances in which I got dumped into the job market, I started out with a sysadmin job with some programming, which I turned into more programming, but it was not a great start, and it was only through unique coincidence (about the only person in the community with serious Lisp Machine and UNIX(TM) experience) and connections that I got my first really good job.
Working for Lisp Machines Inc. (LMI), the other MIT Lisp Machine spinoff, in the early '80s.
They were working with Western Digital, which like everyone else at the time was designing a 68000 based workstation, and conveniently enough, it was based on MIT's Nubus NuMachine: https://en.wikipedia.org/wiki/NuMachine
So they were doing all the normal infrastructure for a high end workstation, and LMI was designing a 4 board Lisp Machine CPU with a Nubus interface that would work in one of their machines, with or without the 68000 processor board running UNIX(TM).
People with a serious UNIX(TM) background were actually harder to find in the community at that time....
In the early '80s... Oh, I didn't realize it was so long ago. I was hoping that your experience with Lisp Machines would enable you to enter in some lucrative but secretive niche sector, that would still be using something similar to Lisp machines now. I know a company has built one internally, but I don't have its name in mind.
Oh well, lucky you for having been exposed to Lisp machines.
We came to this conclusion as well; we decided to do bonuses based on Y/Y revenue growth rather than equity. The bonuses are not capped.
This allows us to:
1) Justly reward our employees to the upside (with cash, delivered semi-anually) if things go according to plan
2) Automatically controls costs if we don't perform as a team
3) Achieve upside fairness across early vs late employees since we can adjust the bonus % as we hire each new person
4) Eliminates oddities due to variations in company valuations where swaths of employees end up underwater due to bad luck of timing
Downsides:
1) the tax treatment of bonuses as income rather than capital gains is nominally worse; but given the complications with options, it probably works out better for all but HUGE equity gains
2) this plan might not work well for a company that will be pre-revenue for many years, but that should be a pretty far outlier case.
All-in-all this allows us to offer the opportunity for employees to earn above-market comp without having hope they get lucky with company growth, market timing, and their timing of joining the company.
I worked at a very successful company that used a similar model: flat 50% of profits paid to employees as bonus. This worked fantastically well in the short term. Now, 10 years later, the company is still very successful but the upside has all shifted to the owners due to departures, renegotiations, and new hires not getting the same terms. So at 3 years it looked very pro-employee but at 10 years it looks very pro-owner.
Would have been really "interesting" if there had been a liquidity event...
The owners stuck around. Why didn't the employees? Serious question... were people fired for having too remunerative packages? Or did people just move on mostly?
I do agree it can be hard to make it fair once a company is big enough or growth slows dramatically, but that's a problem to deal with 3-5 years from now :)
As an investor and/or equity holder, wouldn't you rather see a startup reinvest their profits into their business? Retained earnings are not meant to be paid out as dividends if they can significantly increase shareholder value. This is even more true for startups where you are seeking a "hockey stick growth" model. Especially when you consider what is more important at a given time: profit, revenue, users, market share, uniques, etc. I would argue that it's important for a startup to be transparent around all these issues and help investors understand where you are at. It's unfortunate that startups can only rely on financing rounds to get updated valuations of their business. Coming up with internal valuations based on arbitrary multiples that are not validated by investors could be a slippery slope and thus are not worth considering.
Our bonus plan is based on revenue growth, not profits. Most of your counter-arguments are based on critiques of redirecting profits, which in our case is not what's going on. That said, you are right that generally companies shouldn't pay dividends if they have something better to do with the cash.
However, I'd make the argument that paying above market for top talent is about the best thing a startup can do to increase its likelihood for continued success.
So why did we choose revenue growth instead of profits as the basis for the bonus plan? Profits are easily gamed and frankly rare in startups and would not make for an appropriate metric to base bonuses on for an early-stage company.
I am not sure I follow your points about metrics, transparency, valuations, etc. Those seem like concerns unrelated to the structure for giving employees exposure to our financial upside.
Our #1 metric is revenue growth. That's what we want our team focused on. Not vanity metrics, not profits even. That's a management concern. Our bonus plans cover multi-year terms, and they motivate one to do the right thing in the long-term vs short term. There are no issues with gaming the bonus program; moving $1 of revenue forward/backwards by a few months has no effect. Making an extra $1 now at the expense of $2 next year is not rewarded. "Top management" still has to approve overall direction and operational processes, so it's not like anyone even has the opportunity to game revenue numbers at the expense of operating margin. Besides that, we hire good people and if you can't trust
I will say that for some types of models (eg Twitter) this wouldn't work as it's a free-as-in-beer product until they can start doing advertising. But models like that are quite rare. Though even in those cases there is probably a single vanity metric that is theoretically the main driver of future revenue growth which could be used.
Clearly Sam Altman and a few other people in thread have stopped to think about this. :) Part of what he's telling people here is "look, your competition for the engineers who can help you deliver includes Google and Facebook, your expected value has to be comparable to what they can offer."
Of course, that's still talking about equity, which gets back to the fact that it really is best to treat your equity as little better than a lottery ticket -- something with the possibility of turning into a modest bonus and the remote chance of making you wealthy.
> experienced developers conclude that other options are a better use of their time.
Which I suppose is part of the reason the startup labor pool skews young. Occasionally, though, experienced developers get bored and need new opportunities too.
The labor pool skews young in SV and NYC, possibly because the only way to keep the cost of living down is to have multiple roommates, which does not work for someone with a family. In more suburban places, the age distribution seems in line with an industry that is as new as software is.
It's also a failure of the country's political and economic system, at least if you believe that a series of actions and laws culminating in Sar-Box ended the IPO exit shortly after the turn of the century. Other exits tend to be lots harder (my father had a specialty arranging them, and, hmmm, in the '60s also arranged an IPO...) and tend to leave much less money on the table.
Surely this must reduce the quality of the talent pool available to new startups, as the experienced developers conclude that other options are a better use of their time.