3 ms·
Lets look at this from the other perspective. That of big companies hiring product teams. Educating, finding, hiring, training, cultivating and retaining tale
by rjurney 15y ago
Lets look at this from the other perspective. That of big companies hiring product teams.
Educating, finding, hiring, training, cultivating and retaining talent is extremely difficult. Companies recruit at colleges, where students have worked hard enough on their credentials to gain admittance and even harder getting good grades and joining clubs and participating in activities. They will identify a few good candidates. They will recruit these candidates, along with every other relevant company. Making offer letters, a few will accept. These will be trained for months, then cultivated for years. Now you've got to get these people to work together in teams, and some of those teams will work out and some will not. Most workers will leave in a few years, some even sooner. Some will grow, advance and go on to become leaders. One will go on to become the company's CEO. The funnel chart on this system is daunting, going from millions of students to one CEO.
On the other hand, Y Combinator will take anyone with a convincing video, the 'right look' (see: Moneyball) and attitude, and gung ho fervor. The program teaches these teammates to work together to build and ship experiments quickly and get real people using them, and then to adapt to new data to conduct new experiments. Upon graduation, the teams get more money to continue conducting experiments as they seek their market. Most are acquired in 'acquihires.' Buying a working team bypasses the lengthy process outlined above. Huge value to the acquirer, and contracts/incentives tie the team up for a few years.
Bottom line: these acquihires are profitable for the acquiring companies, the founders, and Y Combinator. Making the acquihire machine turn does not require an idea, a cohesive team, or much of anything except the ability to successfully complete the training Y Combinator provides. Some companies go on to find a market, the rest get the degree, a bonus and a job.
YC training is boot camp. It is the new Harvard, formed to match post-web realities.
Or something.
* Not asserting Y Combinator discriminates, but any time 3 people in a room make a decision without a lot of hard data on the candidates' performance at that task... bias must enter. This is one checkpoint in a new system.
- mkramlich 15y agoAnything that lets us bypass bullshit, makework and proxies-for-things-we-want and instead go directly for substance, realwork, and the-things-we-want, is a good thing and should be done much more of. Also, instead of a "student" throwing $50k at an institution in order to "learn" and gain credentials I'd much rather that the student be paid $50k by an institution and learn and gain a better sort of credentials: building things rather than passing tests. Though I don't agree with exactly how he did it, I suspect this philosophy is similar to the one behind Peter Thiel's recent "20 Under 20" anti-scholarships. And yes, ideally it's better to be acqui-hired than hired. For a software engineer with the right talent mix it's the difference between being paid $300k for 3 years work (which is nothing to sneeze at, certainly, but you can't retire with it), vs being paid $3-30m+ -- if the upside scenario happens, of course. And these aren't even mutually exclusive choices.
- rjurney 15y agoKind of funny that an acquihire entitles you to housing in the Bay Area, isn't it? For a few hundred grand, you can get a house in an outlying area.
- Drbble 15y agoJoining a large tech company today, one can get hired for $1 million on a four year schedule, after incredibly higher odds of payout than a startup. The expected value of income is not the reason to found a startup.
- eli_gottlieb 15y agoHarvard is the new Harvard. The best start-up I know is out of Harvard.