How to Build an Iconic Company – Keith Rabois [audio] | Hacker News Reader

How to Build an Iconic Company – Keith Rabois [audio]

nugget.fm

69 pointscraigcannon42 comments

To his point about finding undiscovered talent, VCs are obsessed with finding people they can underpay via selling them on their “vision” and “impact”. They claim they can’t compete with Google and Facebook on comp, but the truth is they don’t even try to do so by offering fair equity deals. Instead they use rigged instruments like common ISOs with liquidation preference and 90 day exercise windows to intentionally screw early employees over. If they offered fair terms, many talented people would love to be involved in small companies over FAANG, but these types insist on screwing people over and paying them with speeches about “visions.” The vision of these VCs is always enriching themselves.

It’s also absolutely hilarious that his example of finding undiscovered talent at PayPal was by looking at Stanford students. He’s trying to make a point that you should do out of the box thinking on recruiting and his best example is recruiting from the most prestigious university in Silicon Valley. It’s almost difficult to distinguish from satire.

Opendoor was founded by Rabois and has a 3-year-exercise window if you're there for 3 years.

Just to be clear: that's 3 years after leaving the company, they'll still honor your options agreement?

Pinterest began this trend back in 2014-ish, with a 7 year exercise window post-departure. A few companies followed suit, but many more established startups switched to RSUs -- and earlier-stage startups stuck to their 90 day exercise window because it's the 'beaten path' and they didn't want to waste their capital paying lawyers to draft bespoke options agreements.

Yes, AFAIK

There was a very good talk (which I can't find right now) by Ben Horowitz about why companies have to use certain instruments for equity-based compensation. There are a number of legal and 'fairness' (HR and morale) issues which limit what companies can and should do.

edit: found it, starts about 16:30 in https://www.youtube.com/watch?v=uVhTvQXfibU

Not to mention ISOs are far better for employees than NSOs. Moving to RSUs actually reduces the risk and reward for employees as switching over, startups usually cut the grant size substantially (as they are issuing instruments that have intrinsic value at issue rather than no intrinsic value at issue).

tl;dr: Founder shares >>> ISO > NSO >> RSU in terms of risk and reward.

ISOs are not universally better. They have a 24 month time requirement while NSOs only have a 12 month requirement.

I believe selling within the 24 month window is considered a disqualifying disposition which effectively just converts your ISOs into NSOs [1]

[1] https://www.mystockoptions.com/content/what-is-a-disqualifyi...

> There are a number of legal and 'fairness' (HR and morale) issues...

Whatever it was, thank goodness this guy didn't get the memo: https://news.ycombinator.com/item?id=11583480

Startups in some sense are risky financial vehicles aimed at transferring wealth generated by employees to founders and investors.

How much risk do the employees carry? If it is that easy, why don't the employees start a business too?

On an exit, its not uncommon for founders to have 20-50x the payout compared to the earliest employees.

Are you telling me the founders took 20-50x the risk/provided that much more in value?

50X the risk is a tough one, often founders aren't paid, and opportunity cost is just so high today.

If a 30-year old engineer is looking to make a jump, a senior IC role at $BIGCO pays $500-750K/yr starting, with significant refreshers each year, cash bonuses and promotion path to even more. The average time to liquidity for a successful startup is 7 years. Assuming the founders take small or negligible salary for the first 4-5 years then take market salaries, their opportunity cost is easily $5M.

So considering a $5M opportunity cost, and the fact the company wouldn't exist without them, they literally can't just up and quit whenever they want, and are working 100 hour weeks for years on end... yeah, I'd say that's fair.

tl;dr: $5M opportunity cost of guaranteed payout vs $100M maybe sometime in the future doesn't seem insane.

a senior IC role at $BIGCO pays $500-750K/yr starting

At age 30:

This is really only at Facebook and Netflix. Its more like 375-450k at Google/Amazon/MSFT/smaller tech companies.

Facebook and Netflix together definitely employ less than 40k engineers. Of which, no more than 10,000 are ICs with starting comp greater than 500k.

So youre really talking about 10k people between those two companies, throw in the other FAANG, and its less than 40k people.

Of those people, few have the well rounded ability to start a tech company and launch a product end to end, with the sales, product, design, tech etc.

Your post makes no sense and is talking about unicorns that dont really exist.

You don't think someone who's smart, talented, and with diverse enough experience/background to found a startup and lead it to an exit would be capable of a staff engineer role at Facebook? What would you consider a fair point of comparison then?

"would be capable of a staff engineer role at Facebook"

Completely different skill sets. The founder can be scrappy with web programming, knowing a myriad of frameworks and some really good css. He's good at selling, networking, getting people on board with his vision. Maybe his real value comes from his domain knowledge, which lets him succeed where others couldnt.

Its a completely different job. The staff engineer could start a database startup, cofound some highly technical startup. But his highly developed technology skills just arent that valuable in an early stage company. Hes not worth 500k at a startup. He doesnt provide the same value he does at Facebook.

I actually disagree with your characterization of the role at FB, however, what would you consider to be a better way to model opportunity cost?

You can work on compilers, distributed databases and other "hard technical problmens" as an intern. Staff Engineer means something quite different.

Staff Engineer is much more about understanding what to build (including the product implications), how to build it, how to communicate that with everyone else. The kind of skills that would make you a highly effective founding CTO.

If we reduce people to pure levels of drive and motivation, maybe it’s a reasonable comparison.

However, in my experience the types of people who thrive in climbing the corporate ladder (Facebook included as a bureaucratic corporation) are very different from the types of people who seek out building their own successful companies. There’s less overlap between the two personality types than you might expect.

Well said.

It’s important to acknowledge that these highly compensated positions do exist in unique scenarios for extraordinary people. They are attainable for someone with sufficient drive, dedication, and luck.

However, it’s equally important to acknowledge just how rare and unique those positions really are across the industry. $750K compensation isn’t a routine occurrence for engineers just going about their lives.

I’m part of a group mentoring program for CS college students. Some of their compensation expectations are completely out of touch with reality due to comments like this. We encourage everyone to do their research and negotiate, of course, but it’s getting frustrating to watch people distraught because they think their $180K TC offer in SV is a “lowball offer” or who think they’re going to get Netflix level compensation from some random company in their small Midwest home town.

Its mind blowing to run the math on how many people work these positions. For instance, I know for a fact that less than 650 people at UBS make more than 500k in investment banking. Take that number, and all the biggest banks, and you have less than 10k people making more than 500k in investment banking at the big banks. This is miniscule. Yet people talk about these careers like they are viable, or even worth the time and effort. They arent, unless you are top of your class at Harvard, have connections, or are irrational.

Engineering is different in terms of the number of positions, but still, the high IC comp is pretty rare once you run the math

Don't forget that VC investment is convertible debt that the company needs to repay even if it flops.

Lack of access to capital.

Most founders faced the same lack of access too, perhaps getting through is the secret sauce that is so valuable?

He is talking about hiring Stanford students in the year 1999. 20 years ago that was considered out of the box.

> VCs are obsessed with finding people they can underpay via selling them on their “vision” and “impact”. They claim they can’t compete with Google and Facebook on comp, but the truth is they don’t even try to do so by offering fair equity deals

It’s not the VCs, it’s the founders making these calls.

Most VCs have no problems with their companies paying market rate for top talent. It’s usually the founders who think they can extend their runway and minimize their own equity solution by minimizing compensation.

Its often the founders, not the VCs, who try to negotiate for smaller employee equity pools.

> Instead they use rigged instruments like common ISOs with liquidation preference and 90 day exercise windows to intentionally screw early employees over.

I'm so tired of this being repeated as if founders choose ISOs to screw over employees, it's flat wrong. ISOs offer the best tax advantages for employees. The 90-day exercise window is a government-imposed thing. If you want it changed, go talk to them.

I'd much rather give an employee an instrument in which they don't have to worry about any taxes at all until they actually want to exercise than have them sign a document they almost certainly don't understand and receive stock they get immediately taxed for and be confused why the IRS taxed them for stock they can't sell and that might end up being worthless.

There's absolutely no reason companies can't issue options that automatically convert from ISOs to NSOs after the 90-day period, and leave the NSOs on the table for ten years. There is no downside to the employee or, really, to the company – other than that the options, which were pitched to employees as part of comp, aren't clawed back to the company's pool as quickly.

I've observed many startup founders who are disdainful of employees who leave, ever, for any reason, and definitely don't want them to receive proceeds from any of the company's future successes.

One of the reasons is complexity.

Yes, there should be evolution on this front. It's a little lop sided, but not trivially easy to fix either.

These instruments are already very complicated.

Not gonna sugarcoat it - this is weak. Companies don’t offer this solution that’s better because it’s “complex”.

This is like when engineers say something is going to take a long time because “there’s a lot of moving parts”.

The reality, IMO, is that employees do not have a seat at the table when it comes to negotiating ownership shares. And, predictably, they end up with the worst part of the deal.

Complexity is absolutely a reason.

Every time a company offers something 'non standard' - it's a huge legal expense and risk.

Very few startups can afford such things.

Like complexity in code scales exponentially, so too does potential legal outcomes.

There are a lot of bespoke things each company could do in light of specific situations, but it's just not worth it.

You're basically saying hat the VC's have too much comp, and are not handing enough to the FAANGers.

Also - the weird tone about VC's 'enriching themselves' ... when that is 99% of the objective of most employees? People would not work at FAANGS for 1/4 the salary. The money is a 'big deal'.

The power imbalance is probably not the issue. VC's are comped somewhere in the ballpark of 'correct'.

There are zillions of new VC firms every year, capital is cheap - and most of them fail. It is actually competitive.

If there's a systematic bias, it's the immense power of the FAANGS and their ability to hold on to top talent and not always putting them to very good use.

So, is this the same guy who was shouting homophobic slurs at an instructor while studying at Stanford in the 1990s and who was later accused by a male employee of sexual harrassment at Square and fired?

Yes. Since he is the Roger Stone of the PayPal Mafia, he’s not going to go away anytime soon.

Usually it's 50% hard work, and 50% being at the right place at the right time. Nothing is guaranteed.

Sounds like someone with a good SV network is starting the Twitter of podcasts

An investor taking credit for founder's work is probably NOT the way to build an iconic company.

https://twitter.com/rabois/status/1260257254957215749

Your comment is misleading. In no way is he taking credit for the founders’ work. He clearly says he was looking to fund someone in that space.

How is that tweet "taking credit for founder's work"?

Also have you seen this guy's resume? https://en.wikipedia.org/wiki/Keith_Rabois#Business_career Yeah -- he's in no need to take credit for other people's work.

I love taking advice from people like Keith Rabois because you know (unlike 99% of other VCs) he actually has experience building mega-unicorns himself.

I read it differently. It reminds me of some advice on pitching:

“Don’t waste you’re time trying to convince hesitant investors you have a good idea. Find the investors who already know it’s a good idea, and then convince them you’re the team whose going to do it.”

I find that longer content works better for me when trying to understand important concepts. I generally need repetition and context (why is this a valuable concept? how'd you get here?). While it's quicker to digest bite sized audio, I tend to forget the takeaway easily.

Have found these links to be helpful in understanding the concepts here:

* https://startupclass.samaltman.com/courses/lec14/

* https://www.youtube.com/watch?v=9HGRap1cJ3k&feature=emb_titl...

Reply on news.ycombinator.com