Hacker News Reader

Comment by natch | Hacker News Reader

Forget candidate N and the difficulty of discerning nuances. That's just pretending no information is available, so it's a made-up argument. Let's not pretend. Just tell them what the very top candidates had. And if they are a middle candidate and not a bottom one, also tell them what kept them out of the bottom. If they're in the bottom, tell them about the top and the middle. Clues for what is promising and what is not.

We're talking about spending a few minutes sharing words that could positively alter the course of decades of people's lives, if they hear the right thing. Seems like a decent courtesy to pay the applicants, if there's no real reason not to.

View on HN

I wonder how much benefit the good-but-not-accepted groups would receive from a few minutes even up to an hour of time over the next few weeks (maybe after the main interviews are done).

The benefit for YC would be if these groups then chose not to do something like TechStars, but continued to work on their idea and came back in +6 months. Realistically, if someone is rejected by YC, gets into TechStars and does it, they're unlikely to do YC.

Assuming there are ~300 "on the bubble" groups on top of 200 interviewed groups (out of 2-3k total), it would cost YC basically a month of someone's time to follow up with them reasonably, and then maybe 5-10% of a person over the next 6 months.

Wow, that's a really good idea. What I really like about it is that with so many rejected applicants per cycle, you can actually determine statistically the impact of giving feedback on later performance. One can make a small investment of time to try it on a small batch, and if it improves the quality of the whole group, a financial case can be made to spend more time on it.

One group of 500 rejected applicants is sent a sentence or two of feedback by a random YC alum (ideally whoever read the application initially, to account for each reader's biases) 1 month after their interview. Another group of 500 is set aside as the control, and is treated as normal. Is the group of 500 applicants with feedback more likely to reapply? Are they more likely to get an interview or get in the second time around?

I like this idea. This task can also be made simplier by just focusing feedback on the company-related sections of the app. This is where having an experienced perspective (feedback from YC) can really help smart but inexperienced founders.

If you fall into the good-but-not-accepted group, chances are good that you can treat all feedback as an opinion to be considered, not a directive to be obeyed.

The class of things I hate the most (having applied to YC, etc.) is when a bad application masks what is otherwise a good or potentially great team. That's something you can learn not to do, and which can be corrected with some coaching.

1) People fuck up videos frequiently. I've personally been studying "how to do good kickstarter videos" a lot, but "how to do a 2-3 minute Demo Day pitch" and "how to do a 1 minute YC app video" are pretty specific, and while a lot of it should be obvious, it isn't all obvious.

The problem is, doing a good "how to do a YC video" is itself best communicated as video, with good/bad examples, so it's a fair bit of work -- it would probably take a couple people a week to do very well.

2) Sometimes, people "bury the lede" and hide the most salient parts of their project deep inside an application in a non-clear way. Sometimes it isn't even listed. I've talked to friends who want to apply to YC who intentionally couched their whole application in "corporate speak", focused entirely on top-down analysis, etc., rather than actually explaining why the product is good, why they're interested in it, etc.

3) There are a lot of items which are hard to verify, and which lose their predictive value as soon as they're widely known. Similar to the well-known interview questions ("what is your greatest weakness?" being a pathological example), where once they become common knowledge, the only value is knowing who prepared ahead of time vs. who didn't, or who is willing to recite something which sounds convincing vs. actually thinking about it.

I'm afraid the Thiel question "what is something you believe to be true which isn't widely held" is becoming that question today.

X) My personal belief is that something like a YC app should be fractal -- good on the surface if you skim it for 10-30 seconds, but then just as good when you read it closely in any area without greater context. That's a specific style, and not universal, and I don't know if the YC Partnership agrees with me. The applications get reviewed many times (with some rejected at each step), and ultimately if you're on the bubble and the application gets debated internally, you want it to help you just as much as it did when it was initially screened. (Obviously, it's better to just be so amazing that you're well in the "yes" range and not seriously debated...)

Reply on news.ycombinator.com