Greplin (YC W10) Grabs $4 Million From Sequoia For Social Search
techcrunch.com
techcrunch.com
Daniel easily could've packed it up when he lost his cofounder - Shai had to go back to Israel a week or two before Demo Day. Instead, he built a prototype of what would become Greplin (in a week), pitched it to investors at Demo Day, secured funding from Sequoia, and is riding a rocketship.
When PG and YC say that they invest first and foremost in founders, this is a perfect example of what they're talking about. Every startup hits a rough-patch, and some even look like they have no hope at all, but the best founders persevere against crazy odds and will themselves to succeed.
Maybe Daniel isn't "persevering against crazy odds" and "willing himself to success". It's apparent from the story that he's a really fucking good coder. With a reasonably accurate estimation of his own abilities, and with all the support and inspiration around him at YC, it didn't take much perseverance and extraordinary will to put something together. He just had to continue doing what he loves and does best... writing code.
I certainly can't speak for Daniel. I get tired of these descriptions of entrepreneurs that make them sound like cancer survivors. I think founders often have extraordinary skill, and that changes the required personal characteristics considerably.
I think that was Aaron's point. Daniel is an extraordinary person. An ordinary person would probably have just given up given a similar set of circumstances.
But, your classification of people into 2 simple categories is exactly the kind of discussion I like to avoid. From the story, I see a person with skill, confidence, and good odds of success. Aaron's description is one of people with willpower and perseverance in the face of almost impossible odds. These are two completely orthogonal issues. It may be that Daniel is all of the above. But, as a community, we should have more sophisticated dialog on what makes folks successful.
If I were to ever start a company on my own, it would be with the knowledge that I have only half of the traits touted around here. I'm more interested, at this point, in the shortcomings of successful founders and how they compensated.
"Extraordinary skill" or not, the fact is the majority of people would have given up at this point. Daniel instead kept doing back to back investor meetings during the day and coding at night. He's of course benefitted from being smart and part of YC but clearly those can't be sufficient conditions for success, if they were we'd have a 0% failure rate at YC.
Harj, I wish you the best in selecting people for the program. But, I don't know what the moral of the story is for the rest of us. Be determined? Never give up? That's not quite right... because that's not sufficient. I'm just tired of the heroic founder myth.
Not specifically in Daniel's case, mind you (he was already in YC with a good idea that's also rather easy to implement, so this made sense), but I will argue that it doesn't generally make sense to persevere when everything is against you.
Pretty much everyone will say "I stuck with it", and almost all of them will be right - they did stick with it, at least for a while. The thing is, could they have held on for even longer? Put another way, almost no startups fail because the founders go completely bankrupt and don't pay the server bill and the site is shut down by their hosts even has the founders still frantically keep working to stave off death. But most people give up at some point before then - after there's a launch but no traction and no investors for three or six or eighteen months, when they are not literally bankrupt or have not literally exhausted every last possibility, but it seems hopeless and that 100K/year Google offer starts looking a lot more rational.
We were shooting to make the whole system less tightly coupled so that, for example, the machine fetching my GMail data, the machine indexing the data, and the machine displaying the front-end website are all able to survive and do work if any of the others are down.
This allows us to scale horizontally just for our current bottleneck, and to be more robust against individual machine or external service failures. Additionally, it helps us build in much deeper security since each piece of our infrastructure has a small and well defined purpose.
The outcome could be either really good or a privacy nightmare.
Greplin's core service really is tied into reading all of your stuff. That is a privacy nightmare in some regards, but all of this data exists in other services already, so you trust someone else with it. Question is whether you trust Greplin i suppose!
A local version of Greplin would be more interesting to me; I think the Locker project https://github.com/quartzjer/Locker might make that possible someday.