On Twitter and Character Limits

So right at this precise moment of internet time, people are saying that Twitter may raise the character limit on tweets, from 140 to 10,000. This may or may not be true, of course, but it prompted some interesting discussion and thoughts which I feel moved to record.

As an ex-software guy, I feel like I understand this. Twitter, the company, is having some trouble and looking to shake things up. As an engineer, the 140 character limit is weird black magic; the number 140 derives from the SMS standard, which has long since ceased to be an issue on phones, much less the internet. It’s a technical limitation people have inexplicably clung to, as though we’d continued packaging files in 1.44 MB chunks even though CDs existed. (If you understood that you are probably old. I hate to be the one to tell you.)

“Well,” people say, “you don’t have to write long tweets. The people who want to can, and the people who don’t want to can still write short ones! More choice is good, right?”

First of all, hypothetical straw man, the question of whether more choice is always good is somewhat contentious, to say the least. It goes beyond that, though, because that thinking takes into account the needs of one group of stakeholders, call them the writers of Twitter, while ignoring the needs of the readers of Twitter.

(Parenthetical. Obviously many of the writers and readers are in fact the same people. However, it makes sense to talk about them separately, because those people might adjust their behavior separately. i.e. if writing becomes harder, I might write less and read more. The same person belongs to different stakeholder groups depending on her task! Software is hard guys.)

A social network, stripped down to its bare essentials, is a mechanism for connecting these two groups. They form a virtuous cycle, what’s usually called the network effect. The more people who are writing worthwhile things, the more readers will join; the more people who are reading, the more writers will join. The key is that both are essential — if all the readers leave, writers will too. (Probably. See below.)

The question to ask, then, is when you make a change to the software enhancing the experience of one group (by allowing writers to write long posts) are you damaging the experience of another group? In this case, judging by the people who are complaining, the answer appears to yes. The people who are against this change aren’t worried about being forced to write long posts, they don’t want to read them, or more precisely have their feeds cluttered up by them.

The success of Twitter has always been difficult to understand in terms of features (“It’s a social network that doesn’t let you write very much!”) but it becomes easier when you look at the features from a reader’s point of view. The 140 character limit doesn’t serve writers of tweets (it can be annoying!) but it’s essential for trying to make sense of a feed; it forces writers to condense their points into a couple of lines, which can then be scanned easily. Reading Twitter is already often compared to drinking from a firehose, but for all that it’s still easier than the alternatives for trying to vaguely keep up hundreds of people.

Facebook, among its other sins, is guilty of serving the writer at the expense of the reader. The easier it got to update, and to embed media in updates, the more useless the News Feed became, until they were essentially forced to solve the problem by culling the feed algorithmically. (Which ultimately goes to a dark place since you can tweak the algorithm at the behest of advertisers.) Now you can’t do on Facebook what was once its basic purpose — keep up with the things that your friends post in roughly chronological order. With each post potentially occupying a large amount of screen real estate, scrolling through your feed is nightmarish.

(Twitter might truncate posts instead, with a “read more” button. This is equally bad, because it leads to a feed full of snippeted half-posts. We already have this functionality — I’m doing it write now, by blogging and then tweeting the link. Except since it’s not automatic, I actually have to think about what to put into the tweet to get people to click on the link.)

I would actually argue that keeping readers happy is more important than keeping writers happy for the health of a social network, for two reasons.

  • Readers absolutely require writers and a good reading experience in order to use the network. But many writers are pathological narcissists and/or teenagers with infinite spare time, and will write even nobody is reading and the writing experience is terrible. (For evidence of this, see 99% of all webpages in the 90s, mine very much included.)
  • There are more readers than writers. Or, rather: Some people read but don’t write, or don’t write much, and most people read more than they write, so they occupy the reader role more often than the writer role.

I am not even sure this makes business sense, except in the extreme short-term. For Twitter’s perspective, people who buy promoted tweets, a subset of writers, are the customers. Readers are the product. So improving the customer experience by degrading the product is a short-term solution at best. Picture a restaurant lowering prices by using poor-quality food. The ability to write longer posts might please advertisers for a while, but if it drives the reader base away it’s all for nothing. (Every site or publication that uses ads faces this dilemma — how much to degrade the reading experience to favor the advertisers. It’s a fine line.)

In summary, I think this would be a bad idea. The 140 character limit is Twitter’s competitive advantage over other social networks, because of the readability standards it imposes on their writers and the ease of reading that creates. I think they meddle with that at their peril. Remember, once you start down the dark path (that is, taking control of the reading experience away from the users) forever will it dominate your destiny.