How we failed our YC interview | Hacker News Reader

How we failed our YC interview

mr-mariusz.blogspot.com

101 pointsmariusz33168 comments

Honestly, I wouldn't take it seriously at all. Find a paying customer. Make them happy. Find another. If you can't find any, try something else.

We got rejected from YC several times and we are doing quite well despite it.

It really means little, unless you make it a big deal. Just move on. The magic you need to succeed is inside of you, not anyone else.

This really needs to be the top comment.

Despite what you think you saw in that meeting, you really have no idea why they rejected you (see Rashomon[1]).

And while writing this kind of post can be cathartic, you're also leaving behind an artifact that will be there to remind you again in the future.

You're almost better off just writing a draft post to get it off your chest, then deleting a few days later.

[1] http://www.imdb.com/title/tt0042876/?ref_=sr_1

Great advice. We aren't thinking too much of it and are excited to [in?]validate some assumptions. We just thought we'd share our experience and what we learned.

Exactly, we launched three months ago, have revenues(4 digits monthly), a healthy 100% monthly growth rate.

We just left the interview really confused about the reasons why they wanted to interview us in the first place. Since all they really did was arguing our market had 0 potential and our app was not useful, which in itself is totally fine. But why interview us, in the first place, if you don't believe in it?

The fact is, we will never know why they interviewed us and subsequently rejected us. We just need to carry on.

Our goal now is to get ramen profitable(we are really close) and prove them wrong ;)

The interview from their perspective is probably an opportunity to learn more about it and see if you can ease any concerns they have. Another possibility is that they may have liked the team and wanted to see if you would consider working on something else.

Totally agree with this. And it is where we probably dropped the ball.

They probably were hoping you'd prove them wrong during the interview.

Or it might have been the classic "husband-and-wife squabble": one of the partners really liked you (possibly not someone in the interview), then convinced the rest (who disagreed) to give you a shot. The others came in pre-biased to reject you, but willing to be just polite enough to appease your advocate partner.

If you really aced the interview, you'd never know it. They'd still dislike you. But they wouldn't be able to say no, since they knew they'd have to tell your advocate what happened in the interview--so they'd give you a "we'll get back to you." Then, later that day, the interviewers and the advocate have a huge argument on the real merits of your company. Your fate hangs in the balance of that argument.

If you've ever run a garage sale, you'll have seen this all play out in real-time on your lawn. :)

I don't think it works like that at YC. That's more common in larger funds, and usually is just an excuse except at the more dysfunctional ones.

Explicit demand and no competition (cancer treatment) — demonstrate feasibility

Explicit demand and existing competition (fashion ecommerce) — differentiate and show competitive advantages

Implicit demand (farmville-like games) — demonstrate early traction and prove that your product is delivering high-order values (entertainment, happiness, inspiration, meaning...)

No demand — delusion

Love the comment but didn't a great man once say "customers don't know what they want till you show it to them"

That's implicit demand

"Implicit demand" by nature can be used by every entreprenuer for most ideas.

Hence, "show traction".

I guess this is the reason why you have to show traction.

Exactly, if they'd had 15 customers paying $50/month and were adding 5 more each week, that'd be a very different story.

Even if the product at that point was just Matlab-in-the-cloud, those customers could tell them what features they're looking for to help the team make the product unique.

The site having no traction is not the reason they did not get accepted. In the YC application they "explicitly" ask if you have customers. Just a guess here but Im thinking PG and company read their app before they invited them for an interview .

Be careful with how to interpret that. They know their problems, but not the solutions. Ask them about problems that they know they want solved.

I wonder how this would work for something like the first guy trying to sell chewing gum...

> So what will you sell?

> It's gum and you are supposed to chew it.

> What do you mean?

> You place a soft solid in your mouth and you chew it.

> You are delusional.

> But I'll make it colored and flavored.

> Please, seek mental help.

In fact I think chewing gum was uber easy to sell. As a replacement to the toothbrush. Also as a substitute for food when you're hungry.

Chewing gum (as in, a chewy substance you chew for the sake of chewing) was discovered/invented a long time ago, probably before money. It was a naturally-occurring part of the tree. If you mean modern chewing gum, it had to differentiate itself from the previous iteration of chewing gum.

This reminds me of a comedy sketch I heard once which was someone returning from the new world in the 1500's and was pitching the idea of smoking to an investor.

"But why would you do that?"

I found the sketch very funny at the time, but have since found out that the real initial pitch answer was "It cures everything!".

"Anything that harms a man inwardly from his girdle upward might be removed by a moderate use of the herb."

And there lies a startup's life - figuring out what the demand is, and demonstrating feasibility/traction.

How is it better than Octave (http://www.gnu.org/software/octave/), which runs most M-files and is free?

EDIT: Okay, someone posted the website, and it IS Octave (http://www.vectera.com/howitworks). So, all the advantages are from the cloud part. I guess it might be nice for people who have to use multiple machines or want to check a simulation's progress on their phone?

My main experience of MATLAB is fixing pathetic spaghetti code from non-programmers (think of complex "scientific models" with no concept of DRY principles), and wasting hours trying to adjust their MATLAB scripts to do things that are trivial in just about any other high-level language out there. The "Abandon MATLAB" [1] blog has several articles that really resonate with me; it's a great read if you enjoy a good rant.

That said, the problem with MATLAB is not a lack of learning resources. On the contrary, in my experience the MATLAB docs are nearly unsurpassed; all that money pays for something! The problem with MATLAB is that it is primarily used by those with no general programming knowledge or broader experience, who are therefore tolerant of insane defects and defaults out of ignorance that there is a better way to do things. If it were any other way, they would not have added a gigantic Microsoft-esque ribbon interface to the latest iteration of MATLAB 2013.

If I were a Christian, I would thank God daily that MATLAB has a ridiculously high price. The expensive enterprise licensing means its use is limited mainly to a few demographics of deep-pocketed non-programmers (e.g. scientists at large universities), and I can continue to rely on the choice of MATLAB as a major code smell.

[1]: http://abandonmatlab.wordpress.com/

Also amusing: http://www.youtube.com/watch?v=1lBeungEnx4

"A solution is a something. But a something isn’t always a solution."

I like this quote a lot: you failed to get into YC but I think you passed on this. You can build a better mouse trap, but its only helpful in places where there are mice.

Take it as a learning experiance, either adjust what you are doing or now you can approach your next idea with improved perspective and experiance. From what I recall the acceptance rate of companies offered an interview is still only about 25%. I expect there are a few viable companies that don't make it in each batch.

(From someone who interviewed in S12, was rejected, and took considerably longer that you to reach similar conclusions to those you have).

Some unsolicited thoughts:

I do use MATLAB a fair bit in my work (PhD student now). I like it quite a bit and to be honest I haven't really had any big pain points with it, at least compared to some other packages that cost even more than MATLAB does (but probably have fewer users).

I think you'll have to find a new swing on what matlab does. Academics get discounted licenses and seem to be fine to pay for them, and I'm not sure if a $2K license is a big deal for corporate users. Without those two groups, I'm not sure how many people are left. Likewise, I don't think a service to learn how to use MATLAB would have that big of market.

Maybe an interesting and disruptive swing would be to enable the power of matlab-functionality to be abstracted to allow less sophisticated users to use it? Wolfram Alpha is in some ways this, but it could be tied into other services better.

Ehhh, I am not sure that this is truly the issue with your presentation.

If it is truly "we can do X just like FamousApplication(tm), except we are in the browser and we can be profitable charging 10% as much as FamousApplication(tm)", that is pretty compelling. That is not something that is really difficult to sell if you can truly deliver and be profitable at that price point.

I am not sure how it is possible you can replace a computationally expensive client side application like MatLAB with something that runs in the cloud, without discovering order of magnitude more efficient solutions for a wide array of well know mathematical algorithms. If you have discovered such then you can pretty much write your own check, if not there is no way you can scale as a service. Thats what I would think you have to convince potential investors about.

looked at your product, it appears to be quite literally octave in the cloud.

Serious question, How can this scale to tens of thousands of users simultaneously using it? It is entirely possible that this is no where near as performance intensive as I am assuming it to be.

Second question, Whats to stop a competitor from doing the same thing your doing at half the price? It seems like it would be relatively trivial for another company to take input matlab programs from a browser, shell out to octave, and then return the results via the http response, and make it look pretty on the client. There may be a whole lot more going on, but from the outside looking in it appears like there is no moat, any technically decent company could enter this niche and do the same exact thing your doing with minimal effort.

Those two questions you have to answer i think rather than whether or not this is a 'solution' or a 'something'

The product right now looks pretty crude, but it also looks like a very early stage prototype.

If they had say 5 good engineers on this for a year, I suspect it would look very much more compelling.

To do that, you need funding, hence the ycombinator interview.

"I am not sure how it is possible you can replace a computationally expensive client side application like MatLAB with something that runs in the cloud, without discovering order of magnitude more efficient solutions for a wide array of well know mathematical algorithms. If you have discovered such then you can pretty much write your own check, if not there is no way you can scale as a service. Thats what I would think you have to convince potential investors about."

I am an engineer (not CE/CS) who uses Matlab, so I am probably missing something fundamental. However, I am curious why you think server side Matlab requires order of magnitude more efficient solutions to current Matlab. Even as a poor college student I used to SSH into my personal desktop running Matlab and it ran acceptably.

At work I will remote into another system running Matlab without issue. Sure if you are doing heavy plot manipulation it may get a bit laggy depending on the situation, but why would mathematical algorithms need to be rewritten unless you mean graphical?

I am basing it on the assumption that you cant pragmatically do 100 matlab sessions on a single server or VM. While a single Matlab session may run great on a server you are SSHed into, i dont think you can run 100 matlab sessions on a single VM or server without problems. Maybe you can in which case great!

If you cant though, then there is some kinda balancing act that the OP has to optimize, how many matlab sessions can simultaneously be hosted on one of the OPs VMs in the cloud at what cost? Since most of the cloud infrastructure providers seem to charge an arm and a leg for CPU, and I am assuming (maybe incorrectly) that this service will be CPU heavy, as an investor I would have to be convinced that this team has some technique for bringing down these costs...

either they would have to have better algorithms, or perhaps some interesting grid architecture that drastically reduces CPU time and hence costs, or something.

I am making assumptions here, but I think these are standard assumptions, these are the kind of assumptions you have to convince a investor you have a solution for. It may be as simple as saying, we can service 10,000 simultaneous users on 10 VMs, I have no clue what the numbers work out to, but they would have to at least show how this can scale profitably.

"I am basing it on the assumption that you cant pragmatically do 100 matlab sessions on a single server or VM. While a single Matlab session may run great on a server you are SSHed into, i dont think you can run 100 matlab sessions on a single VM or server without problems. Maybe you can in which case great!" "10,000 simultaneous users on 10 VMs"

Thank you for the response. I appreciate the insights but I do not agree with the premise that 100 or 1,000 Matlab sessions per server or VM is crucial to the economics of the business.

I know many engineering firms who are perfectly happy and profitable to buy a $2,000+ workstation, $2,000 Matlab license, and X*$Y,000 toolboxes for a single engineer. That engineer might be in meetings all day. They might be on vacation, they might not be working on a task that requires Matlab. Larger firms sometimes get floating licenses that may be 10X or more that cost and still have less 100% software utilization.

I see utilization economies both in software and hardware by licensing out this resource to small and mid-size firms who do not have constant large demand and even large firms who need a surge in computation.

Hey man, just wanted to put it out there, I am sure you have a much better understanding of the space than I do. I dont even think these are 'deal breakers' i just think they are questions that you have to be able to clearly answer when looking for investors... you may have covered them in depth during your interview, just seemed like your original article was trying to divine a mysterious missing puzzle piece that caused you to miss acceptance at ycombinator... from the little info we had to go on seemed to be some basic normal questions that weren't resolved before worrying about the philosophical something or solution arguement

I sincerely wish you great luck and success!

They could just spin up an EC2 instance when someone connects. It's $0.25/hr for a Large (~ 8gb ram, 2/4 core), so if they're servicing someone for a working day (8 hrs), they're out $2. If the user wants to save that working state, they can image the EC2 instance before turning it off (or reassigning it to a different user).

It would also be a nice feature that if you want to do some heavy calculation, you could spin up something more heavy-duty (a high-cpu xl is only 50c/hr). This same workflow with matlab would require you to go and buy faster hardware.

They could do all of that with some kind of standard matlab clone and probably get customers. If on top of that they had some parallelization secret sauce (some way to spin up new ec2 nodes on demand to do calculations), then they've really got quite a product.

The problem isn't the networking part.

The problem is running multiple CPU bound processes on a single server. Cloud systems are great for stuff requiring storage. It's super easy to make a 1 Tb disk look like one hundred 10 Gb disks. The way files are used and the way storage systems are designed, it's transparent to the user that their 10 Gb "disk" is really a VM's fake hard drive.

But it's not possible to do the same thing with the CPU when any of the users are running CPU bound computations. Performance would be terrible. It's particularly bad for Matlab/Octave because they can take advantage of multi-core CPUs. For a lot of stuff the problem could be alleviated by giving each user their own CPU core, but with Matlab it would mean their code is running 1/8th the speed they could get if they ran it locally.

For what it's worth, I need your product and know other people that do too. I would pay for it, providing I didn't have to pay much.

Increasingly, we have big data sets, and want to work with them collaboratively. Because they are so big, the problem of how to proccess the data is added to the old problem of what to do with it. I think that centralizing the 'how' would be a great selling point too.

Maybe I explained that badly - in Amazon AWS are a bunch of datasets.If I want to play with them, I have to create an AWS instance, with the right tools, go find the data, and spend time figuring out how to read it. If your online Matlab also helped me handle data, then for opensource data someone would only need to solve that problem once (ingest), and for closed source data, at least I'd have a common platform with my team.

Anyway, my imagination of what your product could be, and all the problems it could solve, has me excited. I encourage you not to give up or lose faith in the concept.

This is entirely putting aside that yes, Matlab is expensive, and for occasional / home users, it would be nice to have something subscription based rather than couching up for a perpetual license.

There was a startup featured on HN a few days ago (forget the name) that aims to do similar things. Essentially, they wanted to provide native linear algebra and associated data analysis tools for large data sets by providing the user with a Matlab/R like interface, and running appropriate Hadoop/MapReduce code in the backend. They are backed up by some of the biggest names in the machine learning community, including Tibshirani of EOSL fame.

Good tip. I'll look out for them.

I'd love to chat more with you. Can you send me an e-mail? mariusz.n.lapinski at gmail

To the people who say this is an online Matlab replacement, I do not think the website, http://www.vectera.com/, presents itself that way. First it says "Learn Matlab", to me this presents itself as another online programming class. I am not saying this is a bad thing, I personally use Matlab every day and see the value. However, I could not find anywhere it talked about online Matlab collaboration.

Also even though the word Matlab is all over the front page, when I went to the about section it said it was powered by Octave. Again, this is not an inherently bad thing. However, I thought this was an online time-sharing license of genuine Matlab. I would pay $9/month for that; I would not pay $9/month for Octave which I have installed on my notebook.

Also you can generally get a student version of Matlab for around $100 so if the only purpose is learning and not "production" Matlab, I do not think it is fair to compare to the $2,000 version.

Appreciate the comment. However, our issue was how we pitched our big vision, not what we are now (as you see on the homepage, a Matlab learning platform).

"we focused on replacing Matlab"

Replacement should never be the main goal. Rather, making Matlab IRRELEVANT should have been/be prioritized.

trying to improve matlab by building something on top of octave is putting lipstick on a pig. matlab/octave is a great language for writing quick algorithms but it's biggest flaw is it is not designed for productization of those algorithms.

in other words, focus on the real problem with matlab, it's poor interoperability, deployability, and overall ignorance of good programming language design patterns, not it's lack of being in the cloud or having good tutorials. If you really need matlab, you will suffer through whatever god awful things you need to in order to get it to work because it's the only tool that will accomplish your goal. .m files have the tendency to be both the worst and best code you've ever written: algorithmically brilliant, stylistically abhorrent. The fact that I feel dirty after writing one is a sign there is something missing, and it's not tutorials.

As a recently graduated student... ohmygod I wish I had this in college (only because I don't use matlab daily). The tutorials are lovely and I don't have to haul myself to a lab to run it.

I would have paid out of my own pocket in college to have that convenience.

Why is it hard to see the value in making a product that competes in an existing profitable and developed market space, but does so better/cheaper/more conveniently? Isn't that a significant portion of the startups in YC? E.G. Hipmunk - simply a more user friendly flight search and hotel booking? Mariusz shouldn't have to spell it out for YC in this case.

If the issue is as Mariusz describes, he shouldn't be blaming himself, but rather YC instead for not seeing the obvious market potential in fairer pricing. Either YC missed something obvious or the reason YC turned Mariusz down was for a different reason than stated.

Or it could be that YC knows Mathematica and/or matlab are thinking about launching a cloud computing service. Mathematica definitely is and it will most likely ROFLstomp whatever these guys and other competitors are coming up with. Computer algebra systems and numerical analysis aren't just topics you can sprinkle with "Cloud" dust and have it come out right. Unless these guys have years of experience, YC probably made a prudent decision.

There's obvious market potential in building a nationwide chain of stores and outpricing Costco by 1%, but an angel investor isn't going to be interested in giving you a billion dollars to make it happen. You have to have some "hook" that makes them believe you'll be able to rapidly achieve profitability and be SO much better than the competition that you have an unusually high chance of success.

If the problem you are trying to solve is inaccessibility due to prohibitive cost, you'd have to at least show that there is market at that lower price point.

With regards to matlab, I don't believe the cost to be prohibitive. In most situations where you'd even require what matlab provides, $2k is not that much money. There is also already a very good product that competes with matlab on price - Octave, and its free.

Creating value isn't as simple as just making something cheaper.

"MATLAB in the cloud" sounds a lot like IPython + PiCloud: http://blog.picloud.com/2012/12/23/introducing-the-picloud-n...

Exactly, there is definitely a market in further improving Python as the go-to language for scientific computing. Matlab shines with its toolboxes and Python has a long way to go in that respect but as a language is much better thought trough. When I read "Matlab in the cloud" I also figured they simply meant a Python in the Cloud toolset with the appropriate environments/libraries/IDEs.

Citing Writely/Google Drive as a "successful" example is a fundamental misunderstanding of what success is.

A loss leader does not a business make.

The creator made (presumably) a couple of million, and has changed the way we collaborate on documents, I'd love to have that kind of success.

Your understanding of success is a bit skewed as well.

Wealth creation for a single founder based on the _huge_ edge case that a massive company came and scooped you up is extreme LUCK more than anything else.

Writely was, and Google Drive continues to be, a cost center. Is this thing making money, or losing money? That is the sole indicator of business success.

What gets me annoyed is that we also submitted, didn't get an interview and could have answered those questions easily. We've got some good early validation of the fact that we are filling a need (or would once releasing a V1)...

Hah, I was actually trying to pitch a friend on a Matlab-as-a-Service concept a month or two ago--to say that they were skeptical would be an understatement. Good luck!

Well, Adobe offering their toolkit as a cheaper, monthly cloud offering. That is because lots of people cannot shell out $2k upfront for their tools. You are doing similar thing with Matlab. How does it and why do you think you failed?

YC failed to see the future potential of your business because they are looking for the quickest and best bang for the buck obviously.

If you're passionate about your business model, keep going and ignore naysayers who cannot assist you.

I think there is something to be done with an online mat lab product. Google has it right with Chrome OS, computer software won't be around in a few more decades. Everything will be in the cloud. So will matlab, but maybe it won't be matlab, maybe it will be you guys.

There are plenty of examples of companies that didn't get into YC and still succeeded. Kingmaking can only get you so far.

1. A solution (something that solves a problem) 2. A something (something that you think should exist in this world)

Isn't 1. the same as 2.?

If only the author would elaborate on that point. Oh wait, they did.

TL;DR: A something can exist without being a solution. That's not a bad thing, it's just harder to build a business on.

Snarky!

today, same failure happened to my team. we created a solution about cloud gis which takes geomatics engineering concerns into account. some of the investors asked us "what is the difference between google earth and our project". before the presentation, we were joking about what if they ask us about google earth. it wasn't funny.

Fred Wilson wrote a similar experience with Airbnb [1], he's more sincere about it, detail is very thorough. Regard to this article, I only find personal remorse and popular startup names.

[1] http://www.avc.com/a_vc/2011/03/airbnb.html

Whenever I hear someone talk about having a "solution," I want to ask "what is the problem you are solving"?

Too often people cannot describe the problem, or who has the problem. This does not inspire confidence in their solution.

I found it cool when you guys pitched it to us, I still find it cool today!

Would you pay money for it?

good thing there are plenty of accelerators in the sea!

Awesome job, Mariusz

Reply on news.ycombinator.com