Hacker News new | past | comments | ask | show | jobs | submit login
How we got a Y Combinator interview but blew it (keacher.com)
97 points by teuobk on Oct 24, 2012 | hide | past | favorite | 58 comments



Being rejected from YC is not necessarily the end of the world. I speak from experience, when my company first started out, we tried to get into YC S11, got the interview, but did not make it. We continued along bootstrapped for a while, and ended up raising a couple millions in Series A. We haven't had to touch the money, we're profitable and have ~30 people now.

Point is, PG is one of the smartest guys around; but don't be discouraged if he thinks your idea is terrible, he's not always right.


Did you use a different HN id when you applied? I was curious which group you were part of, but there has never been an application including this HN id.


I don't think my HN id was mentioned in the application, but the co-founders were `progolferyo` and `marialegre`.

I'd be super curious what you think of us now, actually. We struggled for a bit to find the right approach but I think we've come a long long way; even though we still have a long way to go.


I found it. I vaguely remember the interview, but not what we thought. I asked the other YC partners if anyone else does.

As for what I think of you now, at your stage you can use the same test I would: monthly revenue growth rate. If it's over 20% that's impressive.


> don't be discouraged if he thinks your idea is terrible, he's not always right

I made it to interviews last year, thought we did OK on the interviews but were ultimately rejected. Looking back, it was definitely the right choice to reject our idea. Holy cow was it bad.

One of my teammates went on to get accepted for a fantastic idea in the following round, though. I since joined http://counsyl.com and am way happier working here than if I was still working on the terrible idea we had (p.s. we're hiring http://is.gd/counsyl_engineering).


Now I am really curious of who you are...


Check his profile.


Why wouldn't you compare it to Dropbox? It makes total sense. As a photographer wouldn't I just drag my photos to a dropbox folder and it would be backed up? I generate a lot of data but not an incredible amount that it can't be backed up through a consumer internet connection.

Now the problem that I face and would like to solve eventually is video editing storage and retrieval. Ever wondered how raw video gets sent around today? Through FedEx in removable hard drives. The sheer amount of raw data is staggering when making any sort of film. Then storing it for archival. All a pain. 1TB of video is common these days. I have a few ideas to solve this and I think videographers would want a solution.


Digital photography is just about entering the realm of being transportable over the internet. I've enjoyed many trips over the last 5 years and have taken many photos, up to about 1TB worth. Backing that up over my home connection takes about 2 months, which I'd say is just about acceptable for something that doesn't need to be available online immediately, and has already been backed up to a local solution.


Backing up online has two issues: 1) cost. Dropbox does not offer very high storage solutions so far. I have 3TB of videos and photos, where am I going to store it ? 2) access speed and upload speed. Especially upload. It takes forever to upload, unless maybe you have fiber connection...

The technical solution for most people is just not there yet. It's way cheaper to backup on a solid drive and store it at your friend's.


There are a few online backup services which let you mail them a disk if uploading would take too long. A quick bit of googling found Mosaic Archive (http://www.mosaicarchive.com/online-photo-backup/), for example. I haven't used them & don't know anything about them, but their website says they provide this service. I'm sure there are others too.


"Now the problem that I face and would like to solve eventually is video editing storage and retrieval."

There are a few solutions out there that address this issue. But they are in the B2B space, mainly positioned to large advertising agencies and production companies, and priced accordingly. AFAIK, there isn't a fantastic consumer-oriented (or even prosumer-oriented) solution in the market.

I'm actually surprised that Apple hasn't tried to address this problem, as an adjunct to its video editing software and its cloud service. Maybe it's looked into the problem and didn't find it big enough to warrant the sky-high IRR it probably requires of a project. But the idea could make sense for a startup.

I imagine you could position this tool to videographers, sure, but also to the growing number of people trying to turn semi-pro at YouTube and other video sites. If your solution is good enough, Google seems like a likely acquirer; these days it's allocating a lot of resources toward building tools to help its prosumer YouTube producers.


Aside from the other points here, Dropbox is much more expensive to get to the amount of storage most photogs have. On top of that, you have to either alter your workflow to be entirely within the Dropbox folder or have a duplicate set, which is prohibitive on a laptop.


It would be great to have a drop-box like solution, but I am wondering if the scale (@ say, ~100GB+) works, though? Consider this guys experience with the data limits on Comcast: http://chris.pirillo.com/comcast-250gb-cap-avoid-dropbox-or-....


I had this weird idea a few years back when couchdb first came out to slice up A/V into "frames" where each frame was a document, then do document syncs to multiple sites as a transfer mechanism.

Problem? Slow as hell....


I could imagine! Were you base64 encoding the frames?


I've got to give you credit for not being at all defensive. I have a hard time with that, and it's something that I need to get better at. When someone gives you feedback, you need to take it graciously. Either they're right, or you didn't communicate yourself well enough, and that's why they're wrong.


Or they are lying.


Weird that they were comparing it to Dropbox, while to me it sounds like they should've compared it to Backblaze or Crashplan.


They were comparing it to Dropbox because that's the comparison actual users will make. Regardless of whether they want to brand as 'backup' or 'photo storage', people are starting to view those things equally and Dropbox is an EXTREMELY popular backup tool [1] [2] [3].

Dropbox even launched a feature to automatically capture photos from a device [4]. If they can't answer the 'compare to Dropbox' question well, they can't defend the business.

[1] - http://www.labnol.org/software/backup-files-dropbox/18634/ [2] - http://lifehacker.com/5406806/best-online-backup-tool-dropbo... [3] - http://workawesome.com/software/data-backup-and-sync-3-dropb... [4] - http://techcrunch.com/2012/10/12/dropbox-responds-to-tumblrs...


I suspect that the target market wouldn't compare it with Dropbox, which gets expensive if you want lots of space. They were building something for serious photographers, with one of the mentioned pain points being the time it takes to backup a photo library. I understand this — I have about half a terabyte of photos (shot in RAW) and I don't take nearly as many photos as some.

That said, Backblaze works fine for me, although the initial backup did take a while.


I don't know the pricing model they were proposing, but the "it costs $500/year for 500GB on Dropbox" line would have probably been where I started.

For me, as a photographer with around that amount needing to be backed up, I'd want to know what their service offered me over the $59/year for Carbonite (that can also back up just about anything else you want).


Yeah, we were surprised at that comparison as well. Most likely, that means we didn't communicate it well enough.


Perhaps half the reason they kept asking about the business model / trying to compare it to Dropbox was that they were trying to evaluate whether they had a potentially billion dollar + company on their hands, or a niche nice little side business.

It has been discussed in detail in one of PG's essays [1] about how it's only the big wins that matter to them these days and trying to determine whether the company has the chance to become that big is all part of the problem.

The fact that you didn't get into Y Combinator shouldn't have stopped you continuing with your idea. It is a problem that needs a solution and your solution may have ended up being a multi million dollar idea which would have given you both great success. The problem is, the start up space these days and incubators in particular are on the hunt for the billion dollar ideas - thus they can gamble on many, for the few successes to help break even.

At the end of the day, you guys seem really switched on and the effort / level of detail you went through suggests you are capable of success - just potentially you need to re apply with a bigger idea next time, after reading PG's 'Black Swan' essay to understand what they're wanting to hear you pitch.

[1] http://paulgraham.com/swan.html


I think ultimately the problem was centering their 'solution' around offsite backup. Most people don't see it as a solution to a real problem, something pressing, but rather as prevention that they might get around to having. Dropbox, on the other hand, solved from day one the nagging problem of syncing between devices, and went further to solve file sharing among people.


Great post. Kudos for having the resolve to follow through on execution and knowing when to move on. Best of luck on your other project!


So, why would you develop something for photogs but do it on windows instead of mac? that seems like a bad step, right out of the gate


We're both photogs who use Windows. It was a combination of being more familiar with this platform and being a bit myopic about our target audience, since we figured we were a part of it.


For a prototype, it really shouldn't make a difference. People mock up UI's in Photoshop, after all.


That's a mockup, not a prototype. If they're spending development time making a prototype, they should build it for the people who would be using it so they can collect user feedback properly.

I do think, however, that their time would have been better spent fleshing out their business plan/idea and creating some nice mockups rather than trying to slam together a subpar prototype.


So call it what you want.

They got meaningful feedback.


In hindsight, it was a blunder. Fortunately, we were able to port Snaposit to OS X fairly quickly due to our having built it with Python and wxPython.


It seems to me that this isn't the first YC interview story that ended with rejection, and being too set on a certain idea being cited as a possible reason why...


Also, YC can't accept every applicant, even if that application can accept the solution might change as time goes on.

Maybe it's that YC just saw a pitch a few hours/days before about a similar product with a better team/solution. Maybe it's that YC already invested in a few similar companies in the same space. Maybe it's the lack of a track record of the founders in building things. Maybe it's the black shirts. There are a lot of reasons besides being tied to a solution.


Sorry for thread-jacking: I have a strange feeling seeing CC-entry form on blurity.com without any indication who processes payments. Yeah, it's https and all, but I don't know blurity. I know PayPal and maybe a few other payment processors, or there could be some badges of third-party security monitors (McAffee, VeriSign, etc.) that could reassure me that the page is safe.

Just my 2c.


Good point. The payment processor for Blurity is Stripe, but that's not apparent unless you inspect the page's source. I suppose that might explain why some people bail to the PayPal option. Regardless, I'll look into adding something to reassure potential customers of their payment security. Thanks!


On your video, it really is amazing how much better a good mic makes the whole thing -- the video part is kind of irrelevant, but being able to clearly hear you speaking without a ground loop hum, wind noise (people doing videos outdoors for some reason!), fan noise from a computer and built-in mic, etc. is a big improvement over many online videos.


Even I was amazed by how much better the audio became. The camera still recorded audio for sync purposes, and it was remarkable how much more "present" the audio from the Zoom H4n was versus the camera audio. I highly recommend getting some sort of dedicated audio solution for anybody doing online video. Something like the Zoom H1 can be had for less than $100, and it will make a world of difference.


I really like the Blue Snowball ($50-80) USB microphones for a lot of purposes. Not as portable as an H1, but really convenient.

(I love my H4ns, though -- I use them for everything. For conferences I usually have cameras with shotguns on L, board audio on R, and a zoom doing the same thing near the board. I still use a Sony HDV with 3 big chips and uncompressed HDMI out to an Atomos Ninja, though, and HDV tapes for backup. Picking the best audio track and getting it all in sync is a pain, though.)


The problem with this kind of backup scheme for photographers is that the utility in having a backup is huge, but the cost of doing that backup at the bandwidth the photographers need is ludicrous.

We have worked with many designers and photographers, and all of them have the same problem - they can generate hundreds of gigabytes of important resources in almost no time at all, and even with a relatively solid pipe normal backup solutions just don't work.

What they really need is a bulletproof storage device that can be transported to a safe location when it needs to be. LaCie has some products in this space, but they don't offer the service that the customer needs - reliable, accessible storage.

The real solution is tape drives. Try to sell an Ultrium drive to a designer, though.


I don't think photography workflows are really that difficult (even if you shoot raw D-SLR or DMF, it's not THAT huge). Video is where most workflows break down -- you might need 20TB SAN just for your working set, and there's not really any good way to archive ProRes 4:2:2 footage very well. Mostly I see people with numerous 1-2TB hard drives, many of which are infrequently accessed and may go bad.


I'm still looking for an automatic photo archival service. I was hoping that Snaposit was that solution. Good luck in your future endeavors.


Liquor seems to be a common theme in these stories: someone should open a bar outside of YC...


May I ask did PG, Trevor, or Robert have and Ipad or computer in front of them during your demo? -I'd love to know!


Not the OP but from our interview: 6 interviewers, two had laptops, everyone else was working from paper or nothing at all.


Paying for Blurity? Meh. There are free AND open source implementations out there, including SmartDeblur (http://yuzhikov.com/projects.html) that has been recently publicized all over the net. Here's a detailed article for those who are interested in image deconvolution: http://habrahabr.ru/post/147828/


I hate it when people develop software and charge for it. Don't they know there's free, open source versions out there? Who do they think they are, trying to make money?

Yes, there's free, open source versions that do the same as Blurity. It looks like Blurity has spent time developing the user experience as well as the algo's used, even comparing with other tools on the market.


Awwww, how cute your sarcasm is.

Compare the two programs and you'll see the difference in algos. Blurity loses in quality of the end-result. If Blurity is any indication of the quality of the solution that he was pitching, well, then it's no doubt he "blew the interview".


Obvious troll is also obvious.

I have compared them, I've looked at the result images for both that are listed. I'd try them but I really don't care that much.


Every solution has a free and open source implementation.

It's like saying I'll bake a cake at home with the free and open source implementation available online instead of buying it from the store.

It about how much you weigh your time and the quality of your product.

Of course, there are exceptions, the cake you buy from the store might be bad but, that's unlikely if you buy from a reputed store.


I really like the analogy, especially regarding software that needs to run on a server vs an saas option.

Because home baked vs store bought also have some very different qualities, both emotionally and practically. Price is just one difference.

Some people just buy cakes. A few make & buy cakes. People who never buy cakes because they can make it themselves are few, opinionated, and kind of cranky in a lovable way.


There seems to be some confusion about the difference between SmartDeblur and Blurity.

In short, SmartDeblur is a parameterized non-blind deconvolution tool. It handles only a narrow set of blur kernels: focus errors or camera movements in a straight line. Real-world motion blurs seldom conform to SmartDeblur's limitations. Also, the technology in SmartDeblur is nothing new: the commercial FocusMagic has been doing exactly the same thing as SmartDeblur for almost a decade.

On the other hand, Blurity is a blind deconvolution tool. It handles any type of blur kernel, not just when the camera was out of focus or moved in a straight line. From talking with many actual customers, I've learned that almost all real-world motion blur is non-linear in nature. Blurity is well-suited to those situations.

Of course, I encourage people to try both tools and use whatever works best on the blurs in their images.


The first sentence is really not required.


Hmmm, they wanted to create Dropbox for photographers...

Have they heard about Flicker (as far as I remember they offer unlimited space for a few dollars a year)?

What's the big surprise they did not get funding? True, Venture Capital bussiness is totally broken, they throw money into a lot of crap ideas, but not that crap.

I guess that Dropbox for storing MS Office documents or Dropbox for storing MP3 files (I have so many of them!) are also weak ideas, given the generic Dropbox can handle that (and Skydrive or Google Drive too).


Christ your comment is both rude and ignorant. You have no idea what the tech being proposed was. The point is photographers have huge data volumes to deal with. A similar problem exists with media companies who need to transfer around massive video files - and there is indeed a niche indistry that specialise in providing fast file transfers (e.g using UDP not TCP).

Your sarcastic and insulting commentary is not needed here.


The photography industry is worth fortunes, and whilst Flickr does offer 'unlimited' photo storage it's with conditions (maximum filesizes etc).

Dropbox itself can handle photography, it's what a lot of photographers use for sharing shoots back with models but a dedicated one for images only with smart meta data usage and a few other features would make it very attractive to the photographers I know.


Just because you got rejected doesn't mean you should give up or conclude your idea was the wrong solution. Soon all cameras will have wifi capabilities, so my advice would be to consider that if you peruse the idea further.

I was very surprised that your rejection email said "you guys", I know YC keep it casual but that's really stunning word usage coming from an investor. As an EIR I saw many acceptance and rejection emails and never have any been written with such sloppy English. I guess it's part if the YC image, but I'm still surprised.

Your video sucked, and I'm sure you're far better at math and software than presentation skills, but that stuff doesn't matter if the idea is solid. You can hire those people later. Check out the kickstarter video Brydge for an example of how to make a great video.

I would agree with others though, forget Windows, build prototypes for a Mac, because every single photographer uses a Mac.

I think your CSI-like unblurr proves you go both have first class skills, keep at it, there's definitely a product that's not yet been built that deals with photographers' needs specifically. Dropbox isn't is for sure.


Regarding the video, we were instructed not to spend too much time on it. As it was, we spent too much time on it. Making a great video is difficult, and ours was there just to meet the requirements of the application. Since we got the interview, clearly it was good enough.




Join us for AI Startup School this June 16-17 in San Francisco!

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

Search: