Thinking differently about Twitter

There are 2 ways to think about social data:

1. What people are saying
2. What people are reading (watching, listening).

If I ask you to estimate how many people are "active" on Twitter what do you imagine? I tend to imagine “activity” in terms of doing stuff, like posting content or replying with comments.

However, the reality is that a majority of Twitter users don’t post any content. Twitter's own definition of what it means to be active acknowledges that most of their users would most accurately be described as content consumers, and not content creators.

Compared with the readership of a traditional media outlet like The New York Times, Twitter is 23 times larger.

Active users on Twitter approach nearly 200 million on average each day according to the most recent annual report. (37MM average 'monetized Daily Active Users' in the United States and 155MM average mDAU in the rest of the world—i.e., people who open the app and get served ads).

The New York Times’ boasts 8.4 million daily readers.

So what about the content?

Turning to the creators, the platform is indeed a fountain of content. Each day content creators post around 500 million tweets. It's a jaw dropping number by any measure.

But how many of those tweets actually get read by anyone? The answer is that engagement drives visibility.

Think about what you actually see in your own Twitter feed daily. Open the app right now and have a look. It’s incredibly rare for any Tweet to appear in your feed that does not have some engagement: at least one "like," at least one retweet, or one quote/reply. In most cases it's much, much more.

Compare this to a platform like Instagram where even a relatively small number of followers regularly generates double digit engagement. On Twitter most people (people who do not read articles like this) are unlikely to get even a couple likes sometimes.

A very rough estimate, based on Talkwalker data, is that fewer than 5 million posts on Twitter have engagement that is significant enough to appear in anyone's feed.

A few-to-many platform

The majority of the content that people actually see on Twitter is limited to a small subset of the total content on the platform.

In the end less than 1 percent of the content on Twitter is seen by most people.

Strip away all that unread content and you are left with something that looks a lot more like a traditional one-to-many publishing model. Twitter is far more similar to the New York Times than whatever we may imagine it to be. A similar argument could be made about YouTube and Netflix.

"It’s what’s happening!"

Perhaps a more accurate slogan for Twitter than “It’s what’s happening” (yes this is their official yet largely unknown brand tagline) would be “Some people are saying.”

The term “social” in social media here is at worst a misnomer. A more nuanced take is that “social” is too easily (and too often) misinterpreted to mean something it’s not--namely a decentralized information platform.

Obviously, we understand that influencers exist. The point is that influencers aren't some of the game, they are all of it when measuring the impact of a marketing campaign. Engagement isn't a multiplier, it's the entire game.

The takeaway for social analysts (and social listening methodologies) is that Twitter should be looked at alongside traditional media outlets more often. And doing so requires limiting your analysis to tweets with at least 1 like or 1 retweet or 1 reply (if not more) because engagement provides the best heuristic for reach.

Is Twitter data still valuable as a source of consumer opinion; for understanding what regular people think and feel? In short, yes. In the end, Twitter provides 2 distinct data streams with value for researchers:

1. Consumer feedback (i.e., What are people saying?)
2. Opinion leaders (i.e., What subset of content are people actually reading?)

Put another way, the comments in the suggestion box reflect what people think, but the truth is that no one is reading them.

‍