Hacker News Reader

Comment by points | Hacker News Reader

Is there an 'official' list that's better? Seems a useful thing to have.

There used to be a list of launched companies in the YC faq, but there started to be so many that it was always out of date. We have a list ourselves of course, but we couldn't publish that because it's full of stuff companies don't want disclosed. And it would be a lot of work to tease apart the stuff that's secret from the stuff that isn't. Often I'm not sure myself. So making a publicly viewable list of all the startups seems to fall into the category of things that might be interesting to do, but wouldn't actually help the startups much, and would be a significant amount of work.

I think this comment at 3:51 PM EST on 10/17/2010 may be the official death knell of structured data.

A few years ago, YC would have maintained portfolio data in a spreadsheet or database. But in today's world, it's probably in a wiki.

It might be surprising that an uberhacker like PG doesn't have a custom database of YC startups (which would, by definition, make it trivial to export the "non-secret" stuff). That's why I think this comment speaks volumes about how -- even amongst supernerds who think in a systematic, structured way -- a loose affiliation of hypertext documents is the preferred way to capture and record institutional knowledge.

Or maybe it is in a spreadsheet and PG is just lazy. But I doubt it.

spreadsheet != laziness (at least, not necessarily)

I think his point was that if it's in a spreadsheet, it would only take a few minutes to copy the "name" column, excluding all startups where the "launched" column indicates that they're still in stealth.

He wasn't calling using a spreadsheet lazy; he was calling using one and still being unwilling to compile a list of YC startups lazy. Of course, PG is busy enough that I wouldn't blame him anyway (but he does have people working for him to do things like that).

Well, in all geekiness, pg could release the structure of 'the list' without revealing the actual data & put an end to all of this structured vs. unstructured speculation. Just sayin'.

We do have a custom database. What makes it hard to export data is that we have no representation of what info is public. The database knows only that e.g. startup x has raised a series A round, or has been acquired, but it doesn't know whether those things have been announced yet.

Would you be willing to release lagged data? 12 months, for example?

There's no way to do that, except I suppose by retrieving a copy of the data from backups.

I believe the only dataset that YC needs to furnish is the official list of alumni.

Follow-on rounds, exits, and current status can all be provided by the community. But there seems to be no authoritative dataset that shows which startups participated in YC.

But from your comments, I'm guessing that some companies prefer that their mere involvement with YC not be disclosed at all, which makes this a harder task then just periodically exporting the name and URL of every portfolio company.

"Rumors of my death have been greatly exaggerated." - Structured Data

There is a reliable source of YC alumni list from outside of YC, but I hesitate to mention it as it includes founders of startups that are still in stealth. If anyone needs the list PM me.

Although with a bit of thinking I'm sure most people can figure out where to get it from.

Why would it be useful?

Judging YCombinator performance against Techstars, etc.

Although I think at this point it's pretty indisputable that YC has a substantial lead on any of it's competitors.

It is interesting just to see what types of companies they fund.

I thought that as well and then I remembered, they don't fund companies or ideas, they fund teams. These teams, may or may not change the idea the had when they applied.

Reply on news.ycombinator.com