Experiences With a Failing Startup

I’m 21 right now and have just graduated from college. I got interested in startups midway through my sophomore year and started my first startup last summer. It’s failing. This has been my experience.

Let’s start the story, somewhere back in middle school.

Two important points to make:

Point #1

I’m a pretty hardcore reductionist. I see things as their components. When I ask myself the question: “could X be done?”, I immediately proceed to break it down: “What would have to happen for me to accomplish X? Answer: X1, X2…Xn. Could I accomplish X1? Yes. Could I accomplish X2? Yes. Could I accomplish Xn?… Yes.”.

I get frustrated with other people who fail to break things down into their components and ask the right questions. From what I recall, this sort of thinking really began to get serious and pronounced some time in middle school.

Point #2

I was always pretty below average at basketball. I loved playing, but was never good enough to make the school team or anything. After trying out and not making it in 8th grade, I decided that I would make it the next year.

This seemed crazy to most people. I didn’t even make the first cuts in 8th grade. So I wasn’t even a top 30 player in my middle school. Next year the middle schools combined so there’d be twice as much competition. How would I make the team then? However, I thought to myself… “Aside from eating, sleeping, and whatever work I have to do to maintain my grades in school, I’m willing to put in whatever it takes for me to make the team next year. Screw friends, video games, TV etc. All I’ll do is practice basketball. If you compare this training to the training of others, my rate of improvement will be much much higher than others. So much so that I’ll easily make the team.”

And, theoretically, I should be able to be the best player in the world. The best people in the world right now practice and work out multiple hours a day. But I could basically triple their efforts. If I literally just practiced shooting, dribbling, finishing, passing, defense, speed, quickness, explosiveness, strength, coordination, conditioning etc. every moment from the time I wake up to the time I go to sleep (taking a little time to eat in between), I would be working 2-5x harder than the current best players, so I should be better than them in 5-10 years (rough estimates).

Anyway, as you probably could have guessed, I didn’t make the NBA or anything. But I did make my high school team. I worked hard, but not hard enough to be the best in the world. I think I still stand by the logic though. Take a kid in middle school who doesn’t have any physical handicaps, have him wake up, work out, eat, work out, eat, work out, eat, work out, shower, watch film, and go to bed. Have him do this every day for 10 years, and he’ll probably be the best player in the world. I know that things like height, arm length and genetics play a role, but the best players in the world don’t work nearly that hard, so the rate of improvement of that kid should be that much faster than the rate of the current best players such that it overcomes any physical deficits the kid might have.

Anyway, I dreamed of being the kid who does this, so the thinking is pretty ingrained in my head. People at the tops of their fields can be beat. You could work a lot harder than them, improve at a rate much faster than them, and eventually be better than them. Extending this logic from beating people in zero-sum competitions like basketball to creating wealth and innovation, I think that a lot is possible. I’m not a wild-eyed optimist, but I think that people are capable of way more than they think they are.

Figuring out what I want to do

Fast forward to later on in high school. At this point I stopped caring as much about getting good at basketball and started caring about more important things. I knew that I would be going to college soon, and starting a career soon after that. I realized how big these two decisions were. For example, I’ll be spending about half of my waking life in my career. So any marginal improvement in the career I choose is multiplied by the duration of my career (that marginal improvement is experienced today, and the next day, and the next day, and the next day…). So then, there’s a huge payoff for making a better career decision. Enough so that researching the career (and college) I choose is one of the highest yielding activities I could be doing at that time. So as a senior in high school and freshman in college, that’s mostly what I did.

After a decent amount of exploring, it became reasonably clear that the top 2 options were academic research, and entrepreneurship. I didn’t know that much about entrepreneurship, but I knew enough to have a sense that I would succeed at it and that it could make me a lot of money and be fun, which are valuable things. But ultimately I started going down the road of scientific research early in college. I was (and still am) very curious about how the world works. I explored a bunch of fields, but decided on neuroscience. I could write a whole book on why, but it was basically because understanding how the brain works is important and gets into a lot of the fundamental questions that I have about how our minds work. I’ll make a quick note that I was also interested in physics and had read a few popular physics books and was interested in string theory and stuff, but I was put off by all the tedious work it involved. I think I overreacted to that tedious work though.

So… sophomore year I became a neuroscience major and volunteered in a lab that did research on vision. This is what they did: drilled a hole in a monkey’s head, stuck an electrode in it which allowed them to record the activity of a neuron, had the monkey sit in front of a screen and follow some dots around with their eyes, and inferred what the neuron’s function was based on how it fired throughout the process.

This whole process took months. It could even take over a year! It takes a long time to train the monkey. First off you have to train it to behave itself and stuff. And you have to feed it and take care of it. And it takes a long time to train the monkey to look at the right dots on the screen. And then to record from enough neurons and do enough visual tasks. And at the end of all of it, there is rarely some sort of substantial discovery that really alters what we know about how humans process information.

I was interested in big questions, and I came to realize that a career in research wasn’t a good way for me to find the answers. I came to realize that answering these big questions, along with other things I want to do, requires resources. At the same time, I read Paul Graham’s How to Start a Startup and thought to myself… “I could definitely do that!!!”. I came to the conclusion that if I want to maximize the impact I have on the world, startups are the best way to do that. The things I could do with the millions/billions I’d make from startups are many orders of magnitude larger than what I could do if I did research or volunteer work or something. Add to that the facts that startups seemed fun (at least for someone like me) and that I thought I had a good chance at succeeding. The result: I decided that I would proceed to learn all I could about startups, and then start them and invest in them myself.

Learning about startups

I started to spend a lot of time learning about startups. I read all of Paul Graham’s essays, other blogs about startups, books about startups, Hacker News. I read about business, marketing, finance, sales etc. I taught myself HTML, CSS and Javascript, starting taking CS classes, started reading about design. Etc. etc. etc.

This all started during the second half of my sophomore year in college. The summer after my sophomore year, I started College Inside View (at that point it was called Collegeanswerz). The idea behind the site is that student reviews of colleges ask questions that are too broad (eg. What are the academics like?). By braking these questions down into their compenents, you could get a much better sense for what a college is like. See this table for a comparison of the questions I ask to the questions that are currently being asked in student reviews.

Starting the site this early was one of my bigger mistakes. The sequence of events went something like this. I had read a lot about startups the previous year and felt ready to start one (I didn’t know everything, but I knew enough that I could learn the rest as I go). I went into the summer planning to learn to code, and then start the site. I barely knew anything about coding, so I couldn’t tell how much I’d need to know in order to start the site. I started learning HTML, CSS, JS and PHP. I planned on learning the basics/fundamentals of each, and then just learning what I needed to build the site. I figured that I just needed to code a prototype that has pages for each college and a place to answer questions about them, and that this would get me the traction I need to raise money and hire real developers to do the rest.

I ended up with a really shitty prototype at the end of the summer. I planned on spending the next semester trying to get people to answer the questions, but decided not to because the site was pretty ugly.

I wish that I spent that summer learning as much as I could about programming instead of just trying to implement features. The fact that I tried to implement features instead of learning what I was doing caused me to learn a lot more slowly than I could have. The reason I made this mistake was that I didn’t know how much programming I’d really have to know.

The next year I was pretty busy with classes, so I just spent my spare time reading more about startups. I also spent a good amount of time doing other types of reading. I read a lot of books about random stuff, but the most important reading I did that year was the Less Wrong sequences. It was by far the most influential stuff I’ve read in my life.

I went into the summer after my junior year with the same game plan as the previous summer: learn what I need to about coding, and then start College Inside View. At this point I had been exposed to Ruby on Rails and knew that that was what I needed to build my site. So I spent the first couple weeks learning Rails, and then started building my website when I felt ready. Similar to the first time, this lead to me copy-and-pasting code so I could implement my features without understanding what I was doing. I did this because I (somewhat impulsively) thought that this was the fastest way to proceed. I thought I only needed to do a couple more things before I got my site to the point where it could get traction and I could hire real developers to do the rest. However, there was always “one more thing” I needed to do. Again, my not knowing how much coding knowledge I’d need prevented me from spending time learning to really code, and this lead to a lot of time lost to coding struggles. A lot of lost time. I’m genuinely disappointed in myself and embarrassed that I made this mistake. I plan on being more efficient with my future pursuits.

Getting started with College Inside View

So I started developing College Inside View about 6 weeks into the summer after my junior year of college. After another 6 weeks or so, I had a version ready to launch! I was pretty confident that the site would succeed. I had broken the components of success into their parts and thought that it was likely that I achieve each component. Enough so that the overall likelihood was pretty high (>90%).

At this point, I thought the 2 major components were: 1) getting people to answer questions, and 2) getting users once I had content. I was (and still am) very confident that I would be able to get users if I had a critical mass of content (enough content such that a user can visit the site and expect to find enough information about the schools they’re interested in). The real question would be if I could get people to answer questions.

I was pretty sure that I’d be able to do it. The most basic argument was, “my competitors get people to answer questions for free, so I should be able to figure it out too”. More specifically, I thought that there was a solid chance that I’d be able to get people to answer questions for free, but if not, worst case scenario is that I pay them. At most, it’d cost a couple hundred thousand dollars a year to pay for the content I need, and I’d be able to make much more than that in yearly revenue.

Let me start off with my belief that I would be able to get answers for free.

• I had a pretty strong sense that there’s a good amount of college students who like talking about their experiences and giving advice to applicants.
• I figured that people would see how useful the site is and would thus want to contribute.
• I figured that people would be inclined to help out a startup that is doing a good thing. I figured the average college student would think have these inclinations, but I especially figured that the silicon valley types would want to help out. That there’d be a bunch of people on HN and Reddit and different bloggers and people I’d come across that would recognize the value that my site provides and would want to help out by answering questions and helping me get others to answer questions. I figured that these people would share it with their friends.

I started off by messaging everyone I knew from high school (which represents a large variety of colleges). I thought that they’d think what I’m doing is cool and would want to help out. I figured that at least a handful of them would eagerly share it with their friends, letting me reach another couple thousand kids. And that of them, another handful would love it and want to share it more. I figured that there’d be enough inclination to help that a snowball effect would inevitably happen and get me a very solid amount of answers. I wasn’t sure if this would be enough, but it’d at least be a good start.

I was pretty confident that I’d be able to get answers for free, but even if that didn’t work, I figured that worst case scenario, I’d be able to get answers pretty easily with a little incentivization. I had lots of ideas for incentives. Paying straight up is the most straightforward, but I could also work with local businesses, or pay a representative to get kids at his college to answer questions, or have some sort of lottery prizes for people who answered questions. I figured that college students love easy money, and would jump at the opportunity to make easy money doing something they’re already somewhat inclined to do (talking about their college experience).

Getting enough content

To my dismay, my beliefs turned out to be mostly wrong.

• Almost no one from my high school responded to my Facebook messages and texts, so that didn’t work out (I was pretty well-liked in high school, so it’s not because everyone hates me or something).
• Reddit, HN, Quora, bloggers etc. were not interested in my reach outs either. At all.
• I ended up getting a decent amount of people to answer questions for Pitt (the school I was attending), but it was like pulling teeth. I put up posters, posted on Facebook groups, talked to everyone I met etc. etc. and people just wouldn’t answer questions. Whenever I’d meet people, the conversation would reach the point of “So what’s you’re major/what are you doing?” and I’d explain my website and how I need content and would really appreciate it if they could go on and answer some questions. They usually would respond by saying something along the lines of, “Oh that’s really cool! That sounds really useful! I’ll definitely go on and answer some questions later. What was the name of it again?”. Of the people who responded this way, maybe 5% of them proceeded to answer questions. I’d often see them again and ask again, but the response was always “Oh yeah, I got caught up with work but I’ll try to do it this weekend.” followed by them not ever getting around to it.
• I also tried emailing the presidents of student organizations at Pitt, telling them about the site and asking them to forward the email on to their club members. I figured that they’d want to help out, but it didn’t work at all. Almost none of the club presidents responded to my email. I didn’t get any answers either, so they presumably didn’t forward it on. The email addresses of club presidents are listed online, so I tried this at other schools too, but with similarly awful results.
• I tried a representative program where I’d pay a current student at a different college to get people to answer questions. I would pay them based on how many answers they got. I had 3 representatives. One was a random guy from Reddit, and the other 2 were my friends. All of them said they were pretty confident they’d be able to get people to answer questions. Given my experiences, I was skeptical, but I figured that they’d at least get their friends to answer questions. One was in a frat, one was an RA, and the other was a president of a student organization, so I was hoping they’d have better results than I did. The result was that they didn’t get anyone to answer any questions. They just kept stalling, saying they were busy but would get around to it and the end of the week. I ended up waiting over a month (maybe even 2 or more months, I don’t remember) for them, hoping they’d get around to it and it’d be successful. Part of why I waited so long was because I was busy with schoolwork. It was wishful thinking on my part. I was hoping that I wouldn’t have to spend the time thinking though and setting up a new program and that the representative thing would work. But the reason I waited so long was mostly because they kept telling me “just a few more days”.
• After that, I said to myself, “Screw this representative thing. I’m just going to pay people to answer questions. It’s going to cost more this way, but I’m pretty confident that it’ll work”. I started off offering $5/10 for 50 answers and I advertised the offer by posting in subreddits. To my surprise, this didn’t work. I really thought that college kids would jump on the opportunity to make 5/10 bucks answering questions about their school.
• After playing with different amounts of money, I did a pilot program where I paid $10 for 25 answers, or $25 for 50 answers. Surely people would answer questions for a more substantial chunk of money right? I did this for 3 schools, and finally had a little success! I figured that if I could get people to answer questions, it would prove to investors that they could invest $x and get y answers. So at that point, they’d just have to bet on the site working out once it had y answers. I figured that I’d easily be able to raise money because it should be obvious that the site has a good chance of working out once it has enough content.

I was wrong two more times.

1. I didn’t really have an easy time getting people to answer questions with the pilot program. It was mostly a matter of finding answerers who would tell all their friends about it, but if I couldn’t find these power users I struggled.
2. Investors didn’t really care about the pilot program. They just wanted to see traction. This approach frustrates me. I talk about it here and here, but essentially I think that investors take a very formulaic approach to investing instead of actually looking at the company in question, breaking down what has to go right into components, and thinking about how likely each component is.

Getting traction

After finishing the pilot program, I spent a bunch of time learning about the fundraising process, how to pitch, reading Venture Hacks, getting set up on Angel List, applying to startup accelerators, polishing my pitch, and looking for investors. Despite my struggles, I thought I had a decent chance at success moving forward. I figured that my company was worthy enough that I’d eventually be able to find someone who would invest. And even if that didn’t work, I applied to a bunch of accelerators and I figured I had a good chance at getting into at least one of them.

Again, my beliefs turned out to be wrong. Investors and accelerators weren’t interested because of the lack of traction, because I’m a single founder, and because I’m not an expert at anything (coding, design, business…).

So moving forward, my plan became to figure out a way to generate traction. I reread http://platformed.info/, and went with the approach of targeting a smaller market, which makes it easier to achieve a critical mass of content. The idea was to target people applying to Ivy league schools by getting reviews for them because I figured that there’s the most demand for student reviews of the Ivy league schools.

So the summer after I graduated college, I started my second pilot program to get reviews for the Ivy league schools. Surely this would work. I would get content, people would find it useful, I’d get traction, be able to raise money, and then would have success once I get a critical mass of content.

Unfortunately, things didn’t run so smoothly. It was difficult getting people to answer questions, even after I bumped the payment up to $1 per answer. After talking to some of the people who answered questions, I realized that 1) people are very skeptical of online ads (I was advertising the offer on Facebook and Reddit), 2) the site seems new and sort of unprofessional, which ads to the skepticism, and 3) people are less motivated to contribute to a site that doesn’t have users yet, because they don’t feel like they’re helping anyone.

However, I also realized most of my answerers came because they were referred to the site by their friends, not because they clicked Facebook/Reddit ads. So I decided to pay $5 per referral, and this led to a good amount of answers. I have over 4,000 answers for the Ivies so far. However, Harvard and Dartmouth have less than 100 combined, so it isn’t too evenly distributed.

Anyway, my ads ran out, people stopped referring and I have a good amount of content, so I decided that the second pilot program was a partial success, and that I’d start trying to get users. I don’t know how I could find/reach high school students who are applying to Ivy league schools, but I do know how to reach college counselors who advise these students (they list their email addresses publicly). So 2 weeks ago, I started sending out emails to them letting them know what the site does and asking them to sign up if they want to be notified when I get more reviews. I figured that this would work. That I’d get a bunch of sign-ups and I’d be able to use this to raise a seed round. Why wouldn’t a college counselor want to be aware of a free resource that provides more in-depth reviews of colleges?

Unfortunately, they’ve almost all ignored my emails. I sent out a couple thousand emails to all the college counselors I could find at US college prep schools (because those are the schools who send kids to Ivy leagues schools, and who have legit college counselors), and have less than 20 sign-ups. A few days ago I decided to stop waiting for them to respond to my emails, and start calling them up on the phone. I called about 100 so far and mostly left voice mails, but hopefully they’ll call back.

From the few I got to speak with I have gathered that they get requests for this sort of stuff all the time, so they’re skeptical. This combined with the fact that summer is starting and they’re busy with the end of the year means that they aren’t paying too much attention to me. Two of the four counselors I spoke with on the phone didn’t even let me say what makes my site different. They just insisted that they weren’t interested in any other resources and were too busy to talk to me. This sucks from the standpoint of me getting traction with what I have right now, but I’m still confident that if I had a critical mass of content, they’d definitely use my site.

Where I’m at right now

So right now it’s looking like my site will fail. I don’t have money, so I can’t pay for the critical mass of content I need. It seems like I need 1 of 3 things to happen:

1. Figure out how to get content for free/very cheaply.
2. Get traction with the content I have right now.
3. Find someone who is willing to invest right now.

I’ve worked hard at all 3 of these things, and it doesn’t seem like they’re happening. I don’t know what to do. My hope is that I’ll be able to get sign-ups from college counselors if I keep calling them on the phone and persisting. But if that doesn’t work, I think this site will fail. Which is really unfortunate, because I truly believe that it’s something that would be really useful to college applicants.

Where was I wrong?

I like to revisit my past beliefs that proved incorrect so I could learn from them and avoid making mistakes in the future. If anyone thinks I was wrong for believing what I believed at any point in my journey, please explain to me why you think so. azerner3@gmail.com

But remember, you can’t use evidence that was gathered after the fact in your argument, because I didn’t have that evidence at the time. Given what I knew at the time, why was I wrong for believing what I believed?

These are my critiques of myself:

1. I should have found a mentor and talked to them about how investors and startup accelerators make decisions. I shouldn’t have assumed that investors are rational.
2. I should have moved much more quickly though the things that didn’t work. It was tough doing the startup while taking classes and living on my own for the first time and learning everything as I go and trying to be a normal college kid… but still, it was a big mistake.
3. I should have learned to code first. I was so excited to get the site up and the business moving forward that I fell victim to wishful thinking and just assumed that I’d be fine with whatever coding skills I had at the time. This was really stupid of me. I should have developed my design skills as well.

Things I don’t regret:

1. Being a solo founder. Some businesses are labor intensive, but not this one (at least not initially). It’s mostly about getting a critical mass of content, and I don’t see how a second person is much help in that area. I think that College Inside View is something that one person can build a version 1 for, get content, raise money, and expand. Plus, I like working alone and am not risk averse.
2. Trying to get people to answer questions for free at first. It was worth a shot. Given what I knew at the time I think it was very reasonable to assume that it is something people might do.
3. Spending almost $5,000 for the second pilot program (with the Ivy league schools). I don’t have much money, but I still think it was a worthy risk for me because it was very reasonable for me to assume that it would be enough to get college counselors to sign up. Fortunately, I could live with my parents, learn to program, and get a job as a developer if it doesn’t work out, so the risk isn’t too big.

Moving forward

I’m still confident in myself moving forward. I’ve been learning Ruby on Rails more thoroughly from the ground up this summer. I still have a ways to go, but I know I’ll be a good developer if I keep practicing and learning. And I plan on being a good designer too. If my site doesn’t work out, I plan on applying to Hacker School, and getting a job as a developer once I am good enough.

But any job I get is temporary. It’s just a way for me to pay for my living expenses while I work towards a new startup. I’m only 21, and I’m still confident that I will start a startup, make a lot of money, use the money to invest in other startups and make even more money, and then use that money to do a bunch of great things for the world.