Essays as Products
I’ve been studying innovation and new product development for years. As a PhD student, I read a lot of academic material about it, which has value. At the same time though, I’ve always wanted to get my hands dirty. I’ve been inspired by David McCullough’s book “The Wright Brothers,” where I learned of the emphasis placed by the brothers on being proficient airplane operators, not just knowledgeable scholars.
Since I am time constrained (married, 2 children, looking for jobs, and finishing a dissertation from a research intensive university), I can’t devote a great deal of time to becoming a “proficient startup operator.” I have successfully hacked together some little things, like YC Remix and Dreamless, but I’m not able to consistently turn out projects without eating into my main gig.
At some point, I realized that Medium essays are products. They can teach a lot about most of the important aspects of development and marketing. Sure, there are a few really important differences: paying hard cash is different than paying with time, for one example. But the parallels are strong: you have to come up with good ideas and execute them well, you have to get people to ‘pay’ for your product (in time), you have to develop a ‘brand’ as an author, and you have to market your products. Fundamentally, you have to create value in either case.
My Most Successful Essay (by Far)
A while back I wrote an essay containing some ideas about power law distributions. Startup successes seem to be distributed according to a power law (or at least something with very, very fat tails, which is the main thing for the point I’m about to make). Why is this important? Because the traditional notion of chance thinks of outliers as ‘special.’ The outliers had some thing about them that made them unexplainable in the context of, say, a normal distribution (the bell curve). With a power law, outliers are expected (that is to say, they aren’t outliers), as long as the sample size is large enough. The implication of power law distributed startup success is that if you try enough stuff, you’ll get a big win eventually.
Okay, but maybe this is facile. Perhaps the pure power law view and the traditional outlier view are both extremes and the truth is somewhere in between. I haven’t totally resolved this question yet, but if I think of my essays as products, I can add some data to the conversation.
The plot that follows is the number of reads I have for each of 20 essays. The format of this graph takes a bit of explaining: it’s sorted from top to bottom according to the number of reads. The technical name for the blue line is an ‘empirical cumulative distribution function’ (plotted upside down). But in plain english, it is a picture of where all of my essays cluster in terms of readership. The top-left dot, for example, is at the top because it’s the essay with the fewest reads, and that number is 4. The x-axis shows the number of reads, but only for a few essays. The blue line is a curve that is meant to pick up the trend in the points (natural smoothing spline if you care.)
What the plot says is that most of my essays sit pretty close to zero. Not right at zero, but close. Then there are a couple with more substantial numbers: 82, 91, 122, 191, 305. Finally, there is the Big Kahuna: 2103. In other words, one single essay has more reads than all of the other essays combined (almost double, in fact).
What happened with that one? Was it just some stroke of genius? No. I had no idea it would take off like that. I almost didn’t publish it, because I thought it would be sort of a dud. But I did publish it, and then Rand Fishken’s tweet happened.
The essay was about how Google’s downfall could come from something that didn’t at first seem directly related to search, perhaps from a low-tech industry. Rand Fishken is one of the founders of Moz, a prominent SEO firm, and somehow he found the article. He has 383,000 followers, and he read my essay and tweeted about it:
Then my readership exploded. I had stepped on a traffic landmine, by no virtue of my own.
All of this stuff is consistent with my conception of power law distributed success: I didn’t have any idea that this essay would succeed ahead of time (as opposed to having visionary confidence in it, as hindsight often tempts us to think), and it took me around 20 tries to get my first big hit (instead of a billion, as normally distributed success would predict).
Again, I’ll point out that this is just one small dataset to think about the question. I think some people really are visionary and one-in-a-billion. But I’m not one of them, so this is actually encouraging for me. I now have hope that if I try enough stuff, I’ll eventually get a big win. In my Medium microcosm, I’ve proved to myself that it can happen.
Sticky Success
One very interesting lab finding, which I’ve referenced before, is that:
people who had failed before had roughly a 20% chance of success on a new startup, which is the same rate as novices. People who had succeeded before had a 34% chance of success on a new venture.
In other words, this study suggests that success is sticky! But can you take this principle from a lab finding into application? How might it play out?
Well, let’s call my essay about Google a success (I realize that 2000 reads is tiny for some people, but it was a big deal for me.) The question is, did it promote the success of subsequent essays? If so, how?
My answer to the first question is yes. I could show you a statistical test, but I’d rather not, so you’ll just have to take my word for it. As for how, there are three major things that happened.
- I was accepted as a writer for Hacker Noon (100K+ followers)
- I became one of the 100 Top Writers in Innovation on Medium.
- Most importantly, I got a big confidence boost.
The first thing is super useful, because now I can submit articles to Hacker Noon without the editors having to ask for it. It’s a much smoother process. That means I’m more likely to publish in Hacker Noon, which in turn drives a lot more readership to my essays.
The second thing means I am a suggested writer, and when people join Medium or want to follow the Innovation hashtag more closely, they see me as someone to follow. I get one or two new followers a day this way.
The most important thing though, is that I realized that a normal person like me can write something that takes off. In some ways, that fact that you can succeed in unpredictable ways is great, because otherwise you might get discouraged by your low chances.
Experiments
So, I continue to write essays. I can use essays to learn more about products, and marketing. I can test hypotheses in my essays (for e.g., I’ve been wondering of late about reputation piggybacking, which is my name for the use of a well-known concept or entity in a title to boost click-through rates. I’m going to try it some.)
I know it’s not as good as the real thing, but it’s the best I can do, given my constraints, and I’ve learned a few things from it. It keeps experimentation costs super low. I can get from idea to finished product (forget MVP) in an hour.
I hope some of what I learned is useful to you. If so, I’d love to know about it on Twitter or wherever.
