OPINION | MEDIUM

Medium Earnings: Read This If Your Earnings Are Still Below August-Early September Rates

Why am I still earning peanuts?

Disclaimer: The views and opinions expressed in this piece are the author’s own which are not confirmed facts from Medium. The author is sharing personal insights and interpretations of the platform’s payment methods, and there has been no official statement or confirmation from Medium regarding the matters discussed. Readers are encouraged to rely on official communications and updates from Medium for accurate information.

Some of us were hopeful, while many others, like myself, were cynical. The fact that you’re reading this now most likely means that your earnings have not gone back to the August-early September rates, as promised by Medium. Mine certainly hasn’t.

The question that is on everyone’s mind now is how are the rates being calculated?

In the article that introduced the August changes, we were told that there are a few major factors contributing to earnings — engagement (claps, comments, highlights), read ratio, follower bonus, and boost bonus. It was not specified whether reading time was a factor, but I’ve read in the comments that it was, although given less weightage compared to the other factors.

In Buster Benson’s latest article, it wasn’t very clear, but he did point out that there had been a change in the incentives. Reading time now contributes more to earnings, and the boost bonus contributes to a smaller percentage. Let’s remove the boost from the equation for now, for a simpler understanding of the calculation.

As per my previous article, which I will link below, I am quite sure that our earnings are based on a revenue-sharing method because if not, how else would Medium have the money to pay us? From what I understand, since there are no ads, there is no other source of revenue for Medium besides membership subscriptions. Basically, this means that all of our earnings have to come from a percentage of all the members’ subscription fees.

When the August changes were implemented, engagement points provided the most rewards to a writer. According to Medium, this led to fraudulent activities in which certain accounts took money away from the rest of us. Medium assured us that actions were taken to stop the fraudulent activities from the end of September through the first few days of October. Measures that led to a decrease in earnings were implemented to prevent more money from going to the scammers. However, most of us still haven’t seen an increase in earnings since then.

I asked three very important questions in the comment section of Buster’s article, but unfortunately, there was no reply. As there is no official answer from Medium, the content hereafter will be my opinion/speculation only. The three questions that I raised are as follow.

Is there is a redistribution of earnings from some writers to others?

Is the drop in earnings caused by fraud or a change in metrics?

Did the new changes in the metrics happen at the same time Medium discovered or stopped the fraud?

In my opinion, the answers are “yes”, “both but mostly change in metrics”, and “yes”, respectively.

Why do I say so?

Well, since Medium has already found out that a high emphasis on engagement points leads to fraud, it makes sense for them to reduce this. Otherwise, the risk remains the same. The problem is that the main reason most of us earned much more in August-early September is because of the engagement points. So, while Medium wasn’t specifically targeting us, it does feel like it, because when the metrics change to emphasise more on reading time instead of engagement, we lose out.

This is purely my speculation as Medium has not confirmed it, but I believe that after the August changes, Medium did not place a limit on how much a reader’s membership can contribute to a writer. Based on my experience from August-early September, each time a reader read and engaged with my article, I earned about $1. It didn’t matter who the reader was; as long as they read for 30 seconds, clapped, commented, and highlighted, I would earn around $1, or more accurately, between 70 cents — $1.10. We will use $1 as the standard rate in our analysis later for easier calculation. The slight variation might be due to the read ratio and follower bonus. It couldn’t have been a coincidence because I observed this with a few other writer friends, and we all reached the same conclusion.

This is not sustainable because fundamentally, all of our earnings have to come from a percentage of Medium membership fees. As mentioned, the only sustainable payment method would be the revenue-sharing method.

How does the revenue-sharing method work?

The actual percentage has not been disclosed, as far as I know, but to simplify, let’s assume that out of each $5 fee, Medium keeps $2 (which includes Boost nominators’ paychecks of up to $900 per month for each of the 100+ Boost nominators), and we receive $3.

Let’s take one reader, Mary, for example. Mary’s $5 will be separated into $2 for Medium and $3 for all the writers that Mary reads.

From the $3, it will be further broken down into a few segments, namely engagement, read ratio, follower, and boost bonus. To simplify this further, let’s exclude the last three because the primary factor causing the significant difference between August-early September and current earnings is engagement points. So we assume that all $3 will be allocated 100% to engagement. This is, of course, not entirely accurate because we’ve removed the other factors, but it serves as a simple reference since the other factors are relatively insignificant when it comes to understanding what is happening to our earnings right now.

So, rightfully, each time Mary reads an article, claps, comments, and highlights, a percentage of the $3 goes to the writer. This percentage depends on how many articles Mary reads and engages with. If Mary only reads articles by one writer for the whole month, the entire $3 goes to the writer of that one article. If Mary reads articles by two writers, the $3 will be divided based on how much Mary engages with each writer. In other words, the more writers Mary reads from, the less each writer earns from Mary’s $3.

So, what happened during August-early September 2023?

During August to early September, Medium seemed to have (perhaps mistakenly) removed revenue-sharing and implemented a flat rate for engagement. Returning to my observation that each read with full engagement earns around $1, this means that Mary only needs to read more than 3 articles for Medium to incur a loss. If 3 articles would already cost Medium $3, and a reader typically reads more than 3 articles a month, can you imagine the scale of this loss when considering Medium’s entire reader base? This loss becomes even more significant when you factor in the impact of the fraudulent activities, where scammers artificially inflate the number of engagements beyond what regular readers do.

I quote this from Buster’s response to Carlo Zeno

Mid-Sept, on-going: Part of the response to the fraud incident is that we updated our fraud prevention system to catch the loophole that they had exploited. A very high-level description of how this works is to make sure that any particular member’s engagement can’t result in author earnings that greatly exceed the cost of their membership. For example, if a member account reads, claps, follows, highlights, and replies to (hypothetically) 1,000 stories a day for 30 days (more than is generally humanly possible without a script or bot), we are taking measures to make sure that a $5 membership can’t result in hundreds or thousands of dollars in author earnings. The engagements will still be seen by the writer, but those engagements will translate to significantly lower earnings, and the writer will still earn at the standard rate from all other members. We don’t cap author earnings in any way.

Based on my understanding, Medium doesn’t only have to ensure that a $5 membership doesn’t lead to hundreds or thousands of dollars in earnings, but even earning just an additional one cent would result in a loss for Medium, which in our example would be $3.01.

In my opinion, they might not have implemented a limit during August-early September because there were/are supposedly more readers than writers. They might have assumed that this model could cover all writers. However, scammers took advantage of this by creating multiple accounts and reading from each other, resulting in substantial earnings. In response, Medium might have introduced a limit on each reader’s fees (revenue-sharing) as mentioned in my previous example. This approach makes it unfeasible for scammers to exploit the system, as they would only receive a fraction of their earnings by reading from each other.

To illustrate this further, let’s consider a scenario where a flat rate is paid (August-early September rates). Let’s also assume that Mary is a scammer and creates 5 different accounts, so she would have paid $25 in total for membership fees. Mary could simply write an article a day on each account and use the other 4 accounts to read from one another every day. Based on my experience of earning $1 per read, she would have earned $4 for each article on one account, totaling $120 for one month on that account alone. Since she has 5 accounts, she would have earned $600. It’s evident that Medium would incur a loss since Mary only paid $25 and made $600. On the other hand, if revenue-sharing method is in place, she will receive a maximum of 5 times $3, which is $15 from interactions with her own accounts, and this clearly wouldn’t work for her since it is less than the $25 that she paid.

Now, even if Mary isn’t a scammer, the situation would be the same if she has at least four writer friends that she engages with on Medium. Ask yourself, do you have at least four writers that you read regularly? Do you see where I’m going with this? Due to Mary and friends’ $25 membership fees, Medium has to payout at least $600 to them, if each of them writes daily. And naturally, Mary doesn’t just read from four writers — obviously the number will be much higher — so the total payout by Medium would be even more significant.

With the revenue-sharing method in place, it prevents Medium from incurring losses, but it comes at a great cost to writers, like myself, who rely on engagement points. In the example we discussed earlier, Mary’s $3 will need to be distributed among all the writers she reads. In other words, it results in only a few cents for each writer if she reads from many different writers, which is sadly the case for most of us.

So what about the emphasis on reading time?

In my opinion, this shift aims to make it more difficult to manipulate the system because it would be time-consuming. Even if someone tried to set up engagement teams to artificially increase reading times (e.g., by leaving pages open longer to increase their friends’ earnings), this will not lead to a huge increase. This change also serves to reduce fraud for similar reasons. Pre-August rates are based on revenue-sharing with reading time as the main/only factor. This was written in the pre-August MPP terms and conditions, but I am unable to share it now as the wordings had been changed in August.

Does this mean that engagement is penalised? Will your own earnings be impacted if you actively engage with other writers’ articles?

I’ve seen this question asked many times as some writers are afraid to engage for fear of making less money if they engage with others. The answer is no. What you do as a reader does not affect your own earnings as a writer; it only influences the earnings of those you read.

Should we stop clapping, commenting and highlighting to avoid negatively impacting our favourite writers’ earnings?

I know that Adrian CDTPPW and Lu Skerdoo are interested in this question. The short answer is no. Any form of engagement is better than no engagement. However, not all engagement is equal. The more writers you read and engage with, the smaller the amount that will be distributed to each writer. Remember Mary’s $3? If you divide it by a large number of articles that you read, you will get a smaller amount. This is unfortunate, but it’s still better than nothing if you don’t read and engage. So, please clap, comment, and highlight as much as you want on all my articles if you enjoyed them. If you don’t do so, I would lose the points, no matter how low they are.

But someone says that articles with less engagement earn more while those with many claps, comments and highlight are the ones experiencing a drop in earnings. So, what is that all about?

I’ve read a comment from someone that goes along the lines of “those who are whining about their earnings are the ones with a lot of engagement on their articles.” That’s not very nice of them. They do have a point because that’s exactly what’s happening, but one is not directly related to the other. Both the statements “articles with less engagement earn more” and “articles with more engagement earn less” could be true for a writer. However, just because both of these statements are true, it does not mean that one of them causes the other. It’s just the circumstances that make it seem like there is a cause-and-effect relationship. I might write about this in another article as this one is getting too long. But please, make no mistake, engagement is not penalised. As explained above, any engagement is better than none. Of course, don’t engage with content that you dislike because if you do so, you will indirectly reduce the earnings on articles that you actually enjoy. So don’t engage with certain writers just for the sake of engaging because in that case, you will hurt your favourite writers’ earnings.

Did Buster lie in his article?

I’ve come across comments accusing him of lying. However, the answer is no — he wasn’t lying. All of this information in my article can be found by analysing his article, so he was indeed telling the truth and stating facts in his article.

The real question is, why can’t it be communicated in simple English instead of jargon? Also, rather than insisting in the comment section that there has been no change in earnings for writers — of course, that’s technically accurate, as it’s simply a shift from those of us who earn more from engagement points to those who earn more on reading time — why not state it plainly? It would be easier for us to understand, sparing us from dissecting each sentence in the article. The drop in earnings isn’t directly caused by fraud. However, it is a response to fraud and a redistribution of our earnings to other writers, moving from those who earn more from engagement points to those who earn more from reading time. Also, he was truthful when he mentioned that there was no cap on a writer’s earnings. Why? It’s because the cap is on the reader’s fees, instead! Of course, the last part was not mentioned specifically in his article, so that was just my assumption to confirm that he was indeed telling the truth. Although, one might argue that lying by omission is still lying? But let’s give him the benefit of the doubt.

Anyway, there might be a reason for this approach. Perhaps they aimed to avoid upsetting their writers, but given the current discontent, it doesn’t seem to have worked effectively.

What about the so-called earnings adjustment?

A few hours ago, in the Earnings page, there was a “one-time bonus for receiving high member engagement” which has been subsequently renamed as “other earnings, adjustments, and bonuses”. Did a Medium staff type in the wrong description or is “high member engagement bonus” an existing benefit? If it is the latter, will we actually receive such a bonus in the future? Has anyone ever received it in the past? But that’s not the important issue here. As expected, the adjusted amount is next to nothing, given that the metrics changed at the same time the fraudulent activities were stopped.

Final thoughts

Of course, I could be totally wrong, and if someone wishes to correct me, please feel free to do so. As I kept stating throughout this article, all of these are mere speculations and interpretations based on my own analysis of the stats and the information in Buster’s article.

Now, over to you. What do you think? Do you agree or disagree with my findings? Let’s chat in the comments.

Edited to add the latest update from Medium on the fraud issue.

And here is the link to one of my first articles on the whole issue.

If you enjoyed this post and would like to support me, please consider buying me a coffee. Thank you very much! https://ko-fi.com/samtzelin