Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
The secret algorithm one VC firm uses to pick entrepreneurs (gigaom.com)
24 points by iProject on Dec 23, 2012 | hide | past | favorite | 19 comments


> We get the age of cellphone number through a process called cellphone underwriting, which reveals all of the above and more, and which is perfectly legal but secretive enough that I won’t reveal how it works. At the end of the day, we’re looking for entrepreneurs who are young, stable, middle class, and who have the support of family and friends networks — and the age of the cellphone number tells us all those things.

> So, I take the approximate age of a founder’s cellphone number and then during due diligence, I get a stat: The average time of their first phone call in the day. If you and your entrepreneur team are making and taking calls at 6:30 a.m PT, you’re probably talking with people back East and there’s a 50-50 shot you’re making north of a million in revenues.

This smells like complete bullshit. There are no other (relevant) Google results for "cellphone underwriting": https://www.google.com/search?q=cellphone+underwriting (add the quotes yourself, they're being removed by HN's comment code)

GigaOM has really gone to the dogs.


They failed to mention that any entrepreneur must also be born with the gift of flight or at least levitation. I've never read so much bullshit in one place. So if I wanted to change my cellphone company because a new company offered me a better deal, and I wanted (or had) to change my cellphone number because the new one is so much better and too much advertising companies had my previous one any ways. Plus I didn't want to go on my family's "family plan" because it doesn't make sense when I talk to my friends on the cellhone most of the time and with my family on Skype, it means I'm not stable? maybe not. But it means I'm smart.

What a stupid ass "algorithm".


It seems their algorithm has codified age discrimination. It's a shame really that this is a component of their strategy.


Except that their algorithm, overall, seems like a failed strategy.


I'm fairly certain that any half decent PI can fetch a cell account history by a phone number.


This is a common misconception. Ever since the Telephone Records and Privacy Protection Act of 2006[0,1] was passed, it has been illegal to obtain telephone records if you're not a law enforcement or intelligence agency.

Edit: Here[2] is a more comprehensive list of records that private investigators cannot legally obtain: banking and financial records, telephone records, credit information, nationwide criminal records, and a comprehensive individual profile. However, I have no confirmation of the veracity of that website, as I just found it with a quick search.

0: http://en.wikipedia.org/wiki/Telephone_Records_and_Privacy_P...

1: http://www.gpo.gov/fdsys/pkg/BILLS-109hr4709enr/pdf/BILLS-10...

2: http://www.diligentiagroup.com/legal-investigation/5-myths-w...


I think most of definitions of "telephone records" involve call history and other stuff that got HP board in trouble.

Here something like a credit statement showing consistency of payments to a mobile operator would suffice - with number portability being a law some brokers would sell you access to a database telling you which carrier the number belongs to, and how long ago it was recycled.


Well, like I said above, credit information is also not so easy to obtain.

But remember, this is all when you're strictly following the law. There are plenty of PIs who are willing to bend the law.


I didn't see one thing in there about evaluating the business. That said, assuming that this isn't a prank post, perhaps this kind of reasoning is behind the fact that many great ideas flounder for lack of funding while many bad ones do get funding. Perhaps the VC's behind those ventures aren't looking at the business at all - they are just looking at the age of the founders' cell phone numbers.

I do believe that alot of this kind of thing goes on though. I invented some authentication technology back in 2004 and couldn't get anyone to take me seriously. After a former friend of mine with serious credibility in the tech investor community came along and formed a business around it, he was instantly able to raise money and soon thereafter sold it for $64 million (incidentally, he screwed me out of the $13 million I was supposed to get, but that's a story for another day). The point is that great technology often goes nowhere until the "right" people are behind it - which probably does all of us a serious disservice.

Rather than rejecting great technology with what the investor believes are weak founders for arbitrary reasons, they should keep teams of what they consider to be great founders around to help those "weak" teams with great ideas.


This sounds more like augury than algorithm.

Unless they have any stats to show me (which I doubt), I'm concluding that there's nothing scientific or even logical about any of these "secrets".

If you claim your method is based on statistics, show me some statistics. This article is the antithesis of the Moneyball approach.


Came here to say this. Show us results, not something that sounds like it was pulled out of thin air.


I found that terrifically lame, right from the moment when they injected "big data" into Moneyball. Yes there's statistics in Sabermetrics, but I'm pretty sure you can fit all the data in a relatively small amount of storage, like a usb key.


Yep. I wouldn't want a VC firm that didn't understand the difference between data and "big data". Good thing I wouldn't pass their tests.


It's dangerous to make the comparison to Moneyball when none of the factors used in this "algorithm" track cleanly to measurable statistics that are indicative to future success like you have with sports.

There is also nothing in this algorithm that seems to account for scrappiness, determination, ability to take hard hits, and a host of other intangibles that make someone a good entrepreneur. I'm not sure you can measure that algorithmically, other than possibly from past entrepreneurial experience, although that's a story better told by the entrepreneur rather than some arbitrary proxy.

If I ever take VC, I hope it's from a VC that believes in me, my team, and what we're doing, and not just whether I fit into their mold of what they think an entrepreneur should look like.


Agreed, Moneyball works because of the vast amount of data that is collectable in baseball. Even more fluid sports like basketball and football have a hard time applying algorithms to player ratings (though it's beginning to happen).

While I do think there are probably some useful objective ways to measure entrepreneurs, unless someone has enough data points under his or her belt the "Moneyball" approach is never going to give reliable results.


I really need to waste less time reading total shit on HN.


Is this a joke? I seriously can't tell


It's a joke, but it's possible neither GigaOm nor the author is in on it.

It's a pseudonym for Larry Chiang, a MLM/credit score/etc. promoter who I believe has some kind of fraught relationship with YC.


It has to be a joke. They claim they can snoop on your private call records (but only for iPhones?).




Consider applying for YC's Winter 2026 batch! Applications are open till Nov 10

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: