An Analysis of My Failed Medium Articles

Photo of author
Written By Thomas Smith

I started contributing to Medium in late 2019, and I’ve written a lot about my successes. I’ve had a post take off and earn over $650, found success with neuroscience-driven headlines, and become a top writer in Machine Learning, Artificial Intelligence, Food, and more.

Now, let’s have a look at the failures!

Why? For one thing, it’s only fair. If I’m going to talk about the pieces that earn hundreds of dollars, I should also talk about the pieces that went nowhere. Otherwise, it gives the impression that I’m some kind of Medium rockstar.

Spoiler alert: I’m not.

I just contribute content consistently and write about a variety of different topics based on my own experience. As I’ve shared previously and at length, succeeding in today’s content landscape requires generating a lot of content. When you do that effectively, you’ll have successes, either through skill (very occasionally) or luck (much more frequently).

Trying to predict what will work in advance is nearly impossible. You usually don’t find out until you hit publish. By generating a lot of content — and the occasional smashing success — you also generate a lot of pieces that do just OK and some that fail completely.

The complete failures are just part of the process. And it’s worth acknowledging and even celebrating them.

Having pieces that everyone hates (or totally ignores) indicates that you’re remaining experimental, putting enough content out there, and placing yourself in a great position to take advantage of the random times when a piece hits on some user/algorithmic nerve and skyrockets.

By looking at these in more detail, I hope to show that any success I’ve had on Medium isn’t due to some innate understanding of the platform. It’s due to producing lots of content and seeing what works. That means you can do it too.

Produce content consistently, learn from the successes, and embrace the pieces that go nowhere, and you’ll be in great shape to capitalize on those situations where a particular piece touches on something and takes off.

Let’s take a deeper look at some of my 2019 pieces that failed and see what we can learn together.

The Experiments That Fail

Sometimes you create a piece of content and have no idea whether it will work. It’s an experimental leap and you know it.

Sometimes these do really well. Earlier this year, I experimented with bringing back old content from a home automation blog I stopped updating in 2015. The blog was taken down by Russian hackers (yes, really), but all the content had been backed up by my friends at the Internet Archive. I downloaded the old blog posts and reformatted and published them on Medium.

None of these have done super well, but in the aggregate, they’ve produced some mild successes. This post about the Raspberry PI, for example, has brought in about $15. It’s enough to buy a few coffees from a piece that was otherwise languishing in the digital vaults of the IA.

If an old piece about obscure hardware does well, I thought, why not publish some short-form pieces about new hardware? Maybe those will have an audience too, and I can get some revenue and readership from posts that take a few minutes to write.

Turns out, they don’t. I’ve published several posts of this type and they’ve all failed to catch on. A representative example is the piece “Is the Nest Learning Thermostat Worth it?” It shares info that has done very well on another channel (YouTube), but here on Medium, it’s gone nowhere. It received a total of 29 views, earning a mere $0.15.

Another failed experiment was writing brief reviews of Bay Area restaurants. I thought I would repurpose restaurant photos that I took for other projects and build them out into short, visually compelling reviews. Surely these would attract interest (and perhaps curation in the strangely specific “San Francisco” Medium category).

Here’s one about a popular local restaurant and their visually stunning brunch. The total? Two views. And one of those was from Medium’s curators when they read the piece so they could reject it.

While a few similar pieces did get curated, most have languished with less than 50 views. None have earned significant revenue. I haven’t given up on this angle yet, but so far, no dice.

Why experiment with random new types of content like these? Why not find a topic where you’re successful and pound away at it ad nauseam? Shouldn’t I spend all my time writing wildly successful long-form pieces about the specific technical details of AI?

These experiments make sense because sometimes they work. I never thought, for example, that Medium readers would want to read 1,000+ words about a specific brand of analog film. I wrote an experimental piece about Kodak’s Ektar 100 anyway, profiling the film in the kind of excruciating detail only a fellow photographer could appreciate.

Turns out, fellow photographers did appreciate it. The piece has received several hundred views and earns a respectable $10 or so per month. It inspired several more pieces and a new Medium publication about analog photography.

Experiments that fail are part of the bread and butter of entrepreneurship or any creative pursuit.

You should collect these kinds of failures. They’re evidence that you’re trying new directions for your company or brand. And they’re the only way to discover new business lines, uses for your existing content, partnerships, etc.

The trick is to try new things quickly and cheaply. You probably need to test 10+ things before you find one that works. So the key is to make it quick and easy to try out new ideas. I didn’t invest more than a few hours into any of these experiments. So I can easily walk away from (or modify) those that fail and put more time and energy into those that look promising.

Those are the pieces that are supposed to be experiments. But what about the pieces you do put resources into and expect to do well? Does this expectation always pan out? Effort must yield results, right?

The Unpublishable

Nope. I’ve written several pieces that were exhaustively researched, had compelling (or so I thought) premises, and appeared in solid publications, but failed to find an audience.

One example is my piece “The Neuroscience of Deepfakes.” People love to hear about the neuroscience of random things! And my other piece about deepfakes had gathered 17,000+ views and $650+ in revenue. I wrote up a carefully researched piece and published it in The Startup, which has 512k subscribers.

The result? 89 views and less than $0.75 in revenue. Why? I have no idea. Despite following a similar formula to other successful pieces (a technical topic, research, a great publication, etc.) it was a total failure.

It’s a powerful reminder that formulas and rules of thumb are helpful, but ultimately if an audience doesn’t like your piece, it’s going nowhere.

Another similar piece titled “Deepfaking my Grandpa,” about the accuracy of Faceapp’s aging functions did even worse. Again, it had a compelling concept and appeared in The Startup but received only 56 views.

My guess is that Medium’s curators took one look at the title and went “Nah.” Even though the piece went into a lot of technical detail and covered serious topics like using AI in policing, the title made it sound gimmicky. A better title would have been something like “How Accurate is Faceapp’s Aging AI?”.

It’s a reminder of how important headlines can be — choose one that doesn’t properly capture the feel of the piece and you’ve sunk it from the start.

Still, at least those pieces got picked up somewhere. I’ve also written pieces that completely failed to get chosen for a single publication, much less Medium curation. These are the total stinkers that you shop around all over the place with no luck at all.

My piece “14 Technical Disciplines Summarized in a Single Quote” is something I had been working on for years. Every time I learned about a new discipline, I’d make a mental note about the single line that could sum up all its practices. Some were ironic, others poignant and inspired by my own experiences.

I finally wrote these up as a piece and could almost imagine the members of different fields sharing it, or at least writing in angrily to say I’d misrepresented them.

That’s not how it panned out. The piece was rejected by three different publications. Medium’s curators passed on it. Social media shrugged. It’s gotten less than 30 views. Not quite the smashing success I’d anticipated.

Sometimes you just have no idea what an audience will like and what they won’t. This is the challenge of working in a scalable industry, like content, art or PR. If I spend an hour consulting about AI, I know I’ll get paid for an hour. But with content, you can spend five hours on a piece and have it go nowhere.

Of course, the exact opposite can occur. You can spend no time at all on a piece and have it skyrocket. Last month, I wrote up a brief post-mortem (Anatomy of a $650 Medium Post) about another piece that had found success on Medium. It was breezy and conversational, and I expected it to pick up a few hundred views (at most) from my regular readers.

What happened? The post started to trend. It’s received 3,700 views and yesterday alone it picked up 20 hours of member reading time. I would never have anticipated that a post about a successful post would become a successful post!

That’s exactly the kind of thing that can happen in a scalable environment. You can write something that you think will have limited appeal and through some unknown factor, it can take off.

If you want the positive elements of scalability (random posts that do unexpectedly well for reasons you don’t understand), you have to be willing to accept the negatives (great pieces that go nowhere for equally opaque reasons).

The key is to keep your ego out of it. It’s too easy to get angry at the platform (or your readers) when a good story goes nowhere.

Don’t. Failures with no explanation are the cost of doing business in a scalable environment. Sometimes a piece just doesn’t find an audience. Recognize that these failures can be just as random and unconnected as the surprising successes. Consider the piece a sacrifice to the content gods, click the New Story button, and move on.

Failures That Aren’t Failures

In looking at pieces that failed, it’s important to differentiate the true failures from those that were never meant to succeed.

A failure is when you expect a piece to do well — or think that it might, as in an experimental post — and it doesn’t.

Some pieces, though, are never intended to pick up a big audience and that’s fine. These pieces aren’t failures at all — they’re things you wrote for the enjoyment of it to reach a super-specific niche or for other reasons which are entirely your own.

When I published a two-minute piece about different kinds of whitespace characters, I didn’t expect it to be curated or read by anyone. I’m tickled pink by the fact that it’s received 12 views. It won’t earn anything, won’t trend, and won’t make it into a publication beyond the one I started to collect exactly this sort of code-related ephemera. And that’s fine — that’s what I intended.

For these kinds of pieces, set your bar for success at one reader. If someone other than yourself reads the piece, you’ve done well.

These kinds of pieces can be liberating to write. They allow you to explore a topic without caring about whether others want to follow along with you. But once in a while, one of these pieces might spark an idea for something bigger or help you connect with an audience that cares about the same hyper-specific thing that you do (even if that audience is tiny).

Being Van Gogh

I’ll leave you with one final thought on failure.

In his lifetime, guess how many paintings the renowned impressionist Vincent Van Gogh sold? Keep in mind that he was incredibly prolific — in a single decade, he painted over 900 pieces.

The answer? One painting. And that sold just seven months before his death.

Fast forward a century and Van Gogh’s paintings are considered priceless. In 1990, one sold for nearly $150 million.

I’m not saying my piece about Fitbit’s changing business models (16 views and $0.01 in earnings) is on the level of a Van Gogh. It’s unlikely to suddenly surge in popularity a century after my death.

But at the same time, the dynamics which applied to Van Gogh’s works can apply at a much smaller level on Medium and other platforms. Sometimes ideas aren’t bad — they’re just before their time, obscure, or doomed by a bad title, poor formatting, or some other deadly sin not related to the content.

And these pieces of content can — and do — have unexpected resurgences. Medium is known for occasionally unearthing a piece written years earlier and showing it to some people to see if it finds a new life and trends.

Take a look at the most popular pieces on Medium. Usually, they’re written within the last month but you’ll almost always find something written way further back. At the moment, a post about time travel is in the top 10. It was written in November 2017, more than two years ago.

I don’t know if the post did poorly at first (the author would have to weigh in there). But the fact that it can trend to the top of Medium more than two years after it was published is a reminder that even old posts can unexpectedly come back and find a broad audience. Maybe there’s hope for my “14 disciplines” post yet!

In the end, if you’re writing on Medium or working on any creative pursuit, failures are part of the equation.

If you try an experiment and it fails, pat yourself on the back. Failed experiments are evidence that you’re testing new things, and not falling back on your laurels and becoming complacent. Then, dust yourself off and run another experiment. Good ideas that work in practice are hard to come by — in entrepreneurship, writing, PR, and many other fields, you have to kiss a lot of frogs to find that prince.

If a good piece fails to engage an audience — or even find one in the first place — give it a shrug and consider it a cost of doing business. Scalability is the Internet’s secret superpower — it’s what allows a piece you spent 30 minutes writing (or a pitch you sent out as a Hail-Mary) to trend and find a massive audience, rewarding you in spades for the minimal effort you put in. But it’s the same dynamic that can lead a great piece to fail dismally for reasons you can’t grasp.

Again, let it go, consider it a write-off, and keep generating more content.

Make sure you know which failures are really failures. It’s fine to write pieces you don’t expect to find an audience. If one person reads them, they’re not a failure at all — even if their readership and earnings numbers look similar to some of your genuinely failed pieces.

And finally, realize that failure is seldom absolute. Even a piece that went nowhere at first can suddenly trend months — or even years — later. Remember Van Gogh — you can enjoy similar dynamics to his (and keep both your ears intact to boot).

Looking back at failed pieces is a valuable exercise — you might see helpful trends, recognize patterns, and even learn something that contributes to your next smashing success!

Leave a Reply

%d bloggers like this: