Six months ago, I got a scratchy cell phone call from Tyler while he was hiking at Arches National Park:

“We’re going to Mountain View!” he exclaimed.

“What?” I asked, scared I had misheard him.

“Y Combinator said ‘yes’ for our interview!” Tyler clarified.

“Woohoo!”

A week later, we were in Mountain View.  We interviewed in the afternoon, and in the evening we were rejected.  Here’s how we got to that point.

Note that I have not titled this “How to get a Y Combinator interview” — that’s a different problem.  This post is simply about what we did.  If you want a proper guide, I suggest Jason Shen’s Unofficial Y Combinator Guidebook.

Y Combinator Funding Application Summer 2012

Company name:

Snaposit

It started one evening in early February 2012, when Tyler pitched an idea to me: “Backup for photographers.”

Since we both dabbled in photography, that seemed like a real problem, and after some cursory research, we felt we could solve it.

The gist was that photographers generate so much data that it’s impractical to do backups over consumer internet connections. We polled some of our photographer friends: Did they currently have an off-site backup solution? Would they pay to have that problem solved? Encouraged by the responses, we pushed forward with our solution: Snaposit.

What is your company going to make?

Offsite backup service for photographers

We decided that our first deadline for Snaposit would be the Techstars Boulder application. Both Tyler and I lived in Denver, so it was an obvious choice. The early application due date was  February 26th, so we filled out the TS application form, spent too long over-producing an application video (which was viewed precisely zero times by Techstars), and sent it in. Fairly quickly, we got a question back from Nicole Glaros of Techstars: when would a prototype be ready?

It was a good question, because it prompted us to actually build the prototype. In a mad dash, we put something together in just a few days. The result was horribly ugly and barely functional, but with the clock ticking, we sent it along anyway. Unfortunately, it was Windows-only, and Nicole seemed to have a Mac. Oops.

For each founder, please list [various biographical info]:

tghw; Tyler G. Hicks-Wright; 2007, Stanford University, MS, Computer Science (Artificial Intelligence); 2005, Rose-Hulman Institute of Technology, BS, Computer Science & Economics (Double Major); Homepage: tghw.com [portions redacted]

teuobk; Jeff Keacher; 2009, Stanford University, MS, Management Science & Engineering; 2004, Rose-Hulman Institute of Technology, BS, Electrical Engineering; Homepage: keacher.com [portions redacted]

Needless to say, we got rejected by Techstars. We laughed it off, called it nothing more than a practice run, and buckled down so that we’d have a really good prototype to show to Y Combinator.

Why did you pick this idea to work on? Do you have domain expertise in this area? How do you know people need what you’re making?

We are both avid, published photographers, with huge photo libraries that we have not been able to reliably back up off-site. We are friends with and have worked with other professional photographers who have expressed that they have the same problem.

As the summer Y Combinator application deadline loomed, we filled out that application and made another video.

Oh, the video. We spent a lot of time on the video, which we justified by thinking that a crummy video made by people claiming to know photography probably wouldn’t be very credible.  I’m not sure that it mattered in the end, but nobody could argue that our production values weren’t high.

Please tell us about an interesting project, preferably outside of class or work, that two or more of you created together. Include urls if possible.

How about an adventure we had together instead? Like when the two of us were backpacking in Denali National Park, Alaska, where we were nearly killed by hidden waterfalls and again the next day by grizzly bears. But both times we overcame!

In the software realm, we’ve successfully created the Snaposit beta together in under a month, with each of us developing a major component.

We set up good lighting (two 500-watt halogen lamps into two silver umbrellas, one on each side of the camera, plus a large white disc reflector on the table in front of us for fill). We shot with a good camera (Canon 7D in 720i mode). We used a dedicated microphone (Zoom H4n’s built-in mics, just out of frame above us).

We did take after take. We drank whiskey. We debated what to cover and what to leave out.

Finally, after a couple hours of takes, we did the final one: the one that would become the submitted video.  Although parts of the video might seem a bit scripted, it was all improvised. Honestly, we probably would have done more takes had the camera’s battery not died about 10 seconds after the final take concluded.

Was the video kind of cheesy?  Yes.  Were dark shirts a bad idea? Sure. Did it work? Apparently.

The video (posted on Posterous) was our only real tool to gauge interest in our application after submission.  We watched the view counter slowly tick up as the days went buy.  As the views went up, so did our excitement.

Our dream came true; Tyler relayed the good news to me.  We booked flights to San Francisco, celebrated, and buckled down to really refine our demo for the YC partners.

The morning of the interview found us in the upper floor of Red Rock Coffee Shop in Mountain View.  We were banging away on the code, trying to deal with some major performance bugs that had emerged when we switched to a local demo web server.

If you’ve already started working on it, how long have you been working and how many lines of code (if applicable) have you written?

We have been working on the code for about a month. Currently, we have about 3,000 lines of Python in the cross-platform client and another 2,500 lines of Python, HTML, Javascript, and CSS in the website.

About half an hour before our scheduled interview time, we made the short drive from downtown Mountain View to the Y Combinator offices.

I’d never before been somewhere so orange.  Hundreds of young guys — they were almost exclusively male, and most seemed to be in their 20s — milled about the large common area.  Laptops bloomed from tables like daffodils in the spring.

We checked in and went through our demo a few more times.  Several other people tried to chat us up, but we politely declined.  We were focused on the interview, and everything else before then would just be noise.

Finally, the time came.  And went.  And we still were outside of the interview room.

Paul Graham came out and made himself a smoothie.  He needed a short break, he said. He was barefoot.

Please tell us something surprising or amusing that one of you has discovered. (The answer need not be related to your project.)

* USGS topographical maps in Alaska are not detailed enough to show hidden waterfalls. Discovering this fact was nearly catastrophic for us.


* The best cinnamon roll yet discovered is about 75 miles north of Fort Nelson, BC


* eBay generates almost 27% more revenue on Sundays than on any other day.

Since research has shown that judges are more lenient on full stomachs, Tyler and I thought it was great that Paul was eating something.

The interview began.  Paul, along with Trevor Blackwell and Robert Morris, grilled us on our business model over and over.  They seemed to accept our problem as genuine, but they seemed unconvinced that our solution was the right one.  Minutes went by.  I felt that I wasn’t paying enough attention to Robert, but I couldn’t seem to fix it.

More time went by. They kept wanting to compare Snaposit to Dropbox, and we kept trying to tell them that we were solving a different problem. Finally, the 10-minute timer beeped zero — and we hadn’t even shown them our demo.

We raced through the demo of our prototype, and the three YC partners seemed unimpressed.

We were ushered out of the room, and I knew that we had failed to make a good impression.  We should have been better prepared to defend our solution.  At the same time, we shouldn’t have appeared so set on our existing solution and expressed more willingness to adapt (should it prove necessary).

Our heads were swimming as we left the YC building. We decided liquor was in order.

We spent a couple hours at a bar in downtown Mountain View and then met some friends from undergrad at a nearby restaurant.  All the while, we were waiting anxiously for a phone call from YC.  It never came.

Halfway through dinner, we got the rejection email.

Start of the YC rejection email

Tyler and I resolved to continue working on Snaposit, and so we did. The private beta led to the public beta, which gave way to the general launch. However, the YC partners’ concerns proved prescient: we had identified the right problem but the wrong solution. Not enough people wanted Snaposit.

After evaluating the opportunities for pivoting and surveying our non-customers to find out what would make them happy, we decided that there were no realistic paths to salvaging Snaposit. We tried, we failed, and we were going to move on. Snaposit will be dead by the end of the year.

In the meantime, my other photography software tool has taken off. Blurity, a tool for unblurring blurry photos, is doing great.

Tyler is also doing well, with no shortage of projects to keep him busy.

Snaposit didn’t work out, but we had a great experience building it. Interviewing with YC was fantastic, and the partners were even sharper than we gave them credit for. Was it all worth it?  Without hesitation: yes.

(For an alternative treatment of this story, see Tyler’s write-up from May and his later discussion about why Snaposit failed)