What I Learned From Creating a High School Social Network

In October 2013, I launched freshStart, an anonymous social networking website for my high school. The premise was simple. Hunter College High School is a small school that starts in 7th grade. By the time you’re in 9th grade, you already pretty much know your entire class. However, these impressions are based on what people were like in middle school. I don’t think I need to elaborate on why that’s a problem. The solution I came up with was an anonymous social networking site exclusively for Hunter students. You sign up, enter your gender, grade, and genders and grades (9–12) of people you’re interested in talking to. You’re then randomly paired with someone who meets these criteria and given a chat interface, with both parties anonymous unless they choose to reveal themselves. This gives you the opportunity to make a second first impression. In short, it was Omegle for Hunter Students plus an appealing elevator pitch. In the Aaron Sorkin version of the story, I created freshStart because I was dissatisfied with my high school romantic situation. He would not be completely wrong on that one.

When I first started talking to people about my idea for freshStart, the responses were mixed. Some people thought it could be really cool and useful. Some people thought it would just be awkward and creepy. I honestly had no idea what it would be like and I was curious. So being the edgy 17-year-old coder that I was (as opposed to the wise, mature 19-year-old coder that I am today), I hacked together freshStart in a couple of days and decided I would release it. Best case scenario, I make a cool thing that people use and it makes for great college application material. Worst case scenario, it’s a complete failure and everyone forgets about it in a week. I was not interested in starting a company. I just wanted to see what would happen.

The first promo I posted to Facebook for freshStart on August 26, 2013

I posted promos like the one above on Facebook. I told everyone I knew about it. Finally, I picked a date and time and released freshStart to the Hunter community. Much to my satisfaction (and somewhat to my surprise), freshStart was a hit. People were using it. People were talking about it. I accumulated a total of over 500 registered users (out of about 800 9th-12th graders) and over 150,000 messages sent. freshStart became a temporary Hunter phenomenon. It also did make for a pretty good college essay. But most importantly, freshStart was a great learning opportunity for me in product development, high school social dynamics, and everything in between. Here are some of the things I learned:

1. Break your app to build hype.

So it turns out, the code that I hacked together in a weekend or so over the summer was really bad. I was handling messaging in a way that’s too embarrassing to share in this article. I even knew that at the time, but I had never built an application that needed to scale before and I didn’t have confidence that this one would. Three hours after launching freshStart, it crashed. I found out by being bombarded with Facebook messages asking why freshStart was down. My shared web host later alerted me that I had used 5x my allotted daily CPU usage in three hours and my account was suspended. People were clamoring for freshStart. I wound up rebuilding it from scratch (in node.js instead of PHP) and rereleasing it a couple weeks later as freshStart: refreshed. The code was still garbage, but it was garbage that could handle a few more users.

I learned two lessons here. The first was a programming lesson: scalability is important. The second was a marketing lesson. As it turned out, my unintentional three-hour sneak preview of freshStart did a better job of building hype around freshStart than anything else I could have done. When I rereleased it, people were signing up and having conversations even quicker than they had the first time around. freshStart was an application that depended on having lots of people using it at the same time, and so my technical debacle became the perfect marketing ploy.

2. You ≠ your app.

freshStart became a Hunter phenomenon for the brief period of time that it was particularly active. I would frequently overhear people gossiping to each other about freshStart. I made the initial mistake of assuming that any conversation about freshStart was one that I was instantly invited to participate in. Once, I tried to enter one of these conversations. I walked over, said “Hey”, and smiled, as if to announce, Lo! It is I, creator of the app about which you speak. I got a sort of “Oh yeah, you made freshStart” acknowledgement, followed by social signals indicating that I was not welcome in the conversation, so I left.

They were right. I had no more of a place in their high school gossiping than I would have had they been talking about Facebook or Twitter. I did not become the social king of Hunter and that was just fine. What I did instead was make an app that became cooler than I was. It was a strange combination of satisfying and dissatisfying. To use a dumb buzzword, I had disrupted my high school. Unfortunately, you can’t disrupt your way to popularity.

My announcement that we reached 100,000 messages sent, exactly one week after the second launch.

3. Data is juicy.

People right now are very concerned about issues of data and privacy. We voluntarily hand over our data to all sorts of companies who do God-knows-what with it. In this case, 500 high-schoolers voluntarily handed over their data to me. Either they really trusted me as a respected member of the Hunter community or they generally don’t think too much about what data they give to whom. I like to think it was the former, but it probably wasn’t.

I got asked a number of questions about freshStart. One of the most popular ones was, so can you like, read everyone’s conversations? The answer to that one is, well, duh. First of all, it would be pretty silly of me to have all this data and no means of accessing it. Second of all, freshStart was a place for high schoolers to chat anonymously, and so I felt that it needed some form of moderation. I included a feature that let users report a conversation. This sent an email to me saying which user reported which conversation, allowing me to review it and decide whether or not to ban the offending user. However, freshStart users signed up only with their email address and Hunter ID number (a number which I could validate but not easily trace to a particular person). If you wanted to be anonymous to me on freshStart, all you needed was an email address that didn’t have your name in it. Yet funnily enough, the vast majority of registered users had emails that clearly identified them.

The next logical question people asked me was, so do you like, read people’s conversations? Let me answer that question with a question. If you were a high school senior and your schoolmates were handing you transcripts of all their juicy gossip on a silver, digital platter, what would you do? My morals were being put to the test. I quickly decided on some important matters of principle. I would not tell anyone who they were chatting with, although many people asked me to. I would not disclose anything I read to anyone unless I thought someone was in danger, which thankfully didn’t happen. I would not treat anyone differently based on anything I read. Under these morally bulletproof guidelines, I set forth and totally read people’s conversations. You would have too. Besides, it was a social experiment and I needed to see what the results were, which brings me to #4.

4. High schoolers are insecure.

This one doesn’t sound like much of a shocker, but hear me out. The whole reason I came up with freshStart in the first place was because it was the sort of thing I would want to use. It wasn’t that I wanted more friends or had some sort of obvious stain on my reputation. It was that on some level I was insecure about what people thought of me and wanted a chance to remedy that. freshStart would connect you with someone who probably already knew something about you, and then let you have a conversation without that baggage. If you messed up and made a fool of yourself, you moved on to the next conversation — no harm done. If things went well, you could reveal your identity and take it from there. freshStart gave users more control over how their peers perceived them with hardly anything to lose, and that seemed to appeal to people. More members of the eligible Hunter community created accounts than those who didn’t. Of course that only meant that people were curious. The interesting piece comes in when you consider who the most active users of freshStart were. The answer turned out to be everybody. All genders, ages, and rungs on the social ladder were represented relatively uniformly. It wasn’t just me and it wasn’t just people like me. Social insecurity was universal in my high school.

At this point, some of the more regular freshStart users might be a little offended. It’s also important to note that writing about how your product’s user base is fundamentally insecure makes for very poor marketing. But being insecure socially in high school isn’t inherently a bad thing. It keeps us conscious of our behavior and stops us from saying or doing stupid things. Of course that doesn’t mean there aren’t negative consequences (recommended viewing: Mean Girls). I created freshStart in part to combat such ills. I don’t know if it worked or not, but what I do know, and what I would feel confident telling any high school student today is this:

It’s not just you, and I have data to back that up.