Elon Musk, 2018 (Daniel Oberhaus)

Since taking over Twitter last week, Elon Musk has found himself embroiled in multiple controversies, including his clashes with two of the most high-profile users on the site, novelist Stephen King (who has 6.9 million followers) and congresswoman Alexandria Ocasio-Cortez (who has 13.5 million followers. Both King and AOC were expressing their discontent with Musk’s plan to charge users $8 a month (originally $20 a month, before King protested) for the so-called blue check, which verified users at Twitter currently get for free.

On Tuesday, AOC tweeted “Lmao at a billionaire earnestly trying to sell people on the idea that ‘free speech’ is actually a $8/mo subscription plan,” to which Musk replied, “Your feedback is appreciated, now pay $8.” That response got more than a million likes, and sparked a debate across the site over Musk’s plan and over his willingness to alienate current blue checks (of which, full disclosure, I am one) in pursuit of more revenue for the company.

My position on this, which I wrote about this week for Fast Company, is that alienating so-called power users — high-volume, high-follower accounts —is a bad idea. The economic reality of Twitter’s business is that it is highly dependent on the content produced by these power users, who generate most of the traffic on the site. Twitter doesn’t pay these users — of whom AOC is one — anything, but it derives enormous economic value from their work. So nickel-and-diming them by requiring them to pay $8 a month for a small perk is a quintessential pennywise, pound-foolish move.

Others disagree. They think power users are well-compensated for their tweets in the form of the exposure that Twitter gives them, which creates more opportunities to build the brand and market themselves, and therefore it’s totally reasonable for Musk to try to get a little money out of them.

I think that’s wrong, and I think it’s wrong in large part because it rests on a fundamentally mistaken assumption about the way people use Twitter. That assumption, which is often implicit in the way people talk about the site, is that users essentially have a set amount of time they devote to Twitter every day, and, roughly speaking, they’ll fill that amount of time with whatever content the site offers them. On this reading, tweets are largely interchangeable — as long as what’s in your feed is vaguely interesting and amusing, you’ll stick around until you’ve had your fill of tweets for the day, and and then you’ll move on.

Now, if this is true, Elon Musk doesn’t have to worry that much about power users deciding to leave the site or tweeting a lot less in response to his plans (the most important of which is not charging for the blue check, but rather sharply reducing the amount of content moderation on the site). If high-volume, high-follower accounts leave, other people will take their place, and it won’t really affect how much time casual users spend on the site.

The problem is that all these assumptions are wrong. There isn’t a set amount of time that people will spend on Twitter — as anyone who uses the site regularly knows — and tweets are not largely interchangeable. On the contrary, tweets vary dramatically in terms of their effectiveness to hook and engage readers, and tweeters vary dramatically in their ability to write effective, compelling tweets.

To put it in the most banal way possible, some people are much better at tweeting than others. The more of those people Twitter has on the site, the more time users will spend on the site, and the higher its traffic will be. The fewer of them it has, the less time people will spend, and the lower its traffic will be.

This seems like it should be an obvious point. After all, the goal of any ad-driven content business isn’t just to add more viewers or listeners, but also to expand the amount of time they spend on the site. And the way to do that is to offer better — as in more engaging, more compelling — content. All of that is as true of Twitter as it is of a radio station or a TV network.

Because Twitter is a social media platform, people tend to think of it as a many-to-many business, in which the value is created by the tweets of its many millions of users. But the reality of Twitter economically is that it’s much more of a few-to-many business, with a small percentage of users producing the content that most other users read and react to. And if the the number and if the quality of those producing the content drops, so too will people’s interest in the site.

In fact, this can happen even if power users don’t actually leave the site, but instead simply spend less time tweeting, either out of frustration or because the user experience has become less enjoyable. Fewer tweets, and fewer tweets that people have invested real time in writing, will mean less interest and engagement from users, and therefore less traffic.

It is, of course, totally possible that the dopamine hits you get from tweeting will prove irresistible, and that there will be no decline in the volume of power users’ tweets, let alone an exodus of them from the site. That remains to be seen. My point is simply that such a decline, or such an exodus, would have a big impact on Twitter’s business, and it would be a bad idea for Musk to make policy in the belief that it wouldn’t.