Twitter still works, in spite of Twitter
This week, Twitter rolled out a website redesign that’s apparently been a year in the making. It features ginormous buttons, frenzied efforts to get you to look at trending hash tags through both “Explore” and “Trends”, and the ability to view up to as many as three tweets at a time.
To no one’s surprise, everyone (including me) hates it.
Since most of the sites we know and love/hate go through major redesigns pretty frequently (Facebook, Dropbox, etc.), a couple of arguments have developed as a way of analyzing these designs:
The users grumble about redesigns, but then go back to using the site anyway and adapt. Just check out any engagement and user base signup figures for Facebook or Twitter over the past 5 years.
The users are not really the users and therefore how they use the product only matters up to the point where their content is helping to sell advertising placements
The product is free and, and the user doesn’t control the platform (always control your own platform), so users don’t really have a cause to complain about any changes
All of these are true, but they’re also dismissive of the real issue. The real issue, argued eloquently by Paul Ford in his must-read essay, “Why Wasn’t I Consulted” (please read this essay if you do any work with the internet in any capacity), is that social media end users put a lot of time and energy into these platforms.
They may be free, but it’s costing them mental points, emotional labor, and all of that good invisible stuff that can’t be quantified in economic terms, to participate, and, as a result, they want the ability to offer input.
Paul writes,
Brace yourself for the initial angry wave of criticism: How dare you, I hate it, it's ugly, you're stupid. The Internet runs on knee-jerk reactions. People will test your work against their pet theories: It is not free, and thus has no value; it lacks community features; I can't believe you don't use dotcaps, lampsheets, or pixel scrims; it is not written in Rusp or Erskell; my cat is displeased. The ultimate question lurks beneath these curses: why wasn't I consulted?
…
That is the point that I am trying to make. The web is not, despite the desires of so many, a publishing medium. The web is a customer service medium. “Intense moderation” in a customer service medium is what “editing” was for publishing.
For Twitter to be really customer-facing, where customers are defined as its users, it would have to elevate the things it’s already good at (from my point of view):
facilitating real-life meetings between people in similar fields or with similar interests (I have legitimately met no less than 30-40 people from Twitter that I never would have intersected with otherwise)
Tracking real-time events, more specifically, conferences
Making weird jokes that only make sense in the context of Twitter:
Functioning like a much better LinkedIn - at least in tech, the amount of people I know who have gotten jobs by connecting directly with hiring managers on Twitter is phenomenal. Twitter is not taking advantage of this.
Creating communities (political, academic, industry, and otherwise)
However! All of these things are incredibly hard to monetize, to machine learn, and to plan a website around. They involve a lot of forethought, and a lot of smart people, both from industry and academia, to analyze how things would work in anticipation. And if there’s one thing that humans are terrible at, it’s anticipating how things will play out.
Just take the Retweet button, for example:
Before Wetherell joined Twitter, people had to manually retweet each other — copying text, pasting it into a new compose window, typing “RT” and the original tweeter’s handle, and hitting send. With the retweet button, Twitter wanted to build this behavior into its product — a standard practice in tech that, at the time, was performed without much thought.
“Quote retweet allows for the dunk. It’s the dunk mechanism.”
“Only two or three times did someone ask a broader and more interesting social question, which was, ‘What is getting shared?’” Wetherell said. “That almost never came up.”
It’s a very hard sociological question - what happens when you change given UI elements? It’s almost never one that gets asked in business meetings.
In Paul’s article, he mentions MetaFilter (which I also love) as an example of a community that has gotten this right. But MetaFilter was set up and carefully curated by humans, and, is now, unfortunately, having serious funding issues. (Insert Normcore Good things don’t scale mantra here.)
Twitter doesn’t want to have funding issues, or to have sociologists sit in on C-suite meetings. Twitter just wants to keep growing and growing. In fact, if you, instead of reading the blog post announcement about how Exciting the new Twitter is and look further, at its 10k form, (which I did last night as I was pumping milk at 3:45 in the morning 🐣), you’ll see what Twitter really cares about, front and center.
mDAUs, in Silicon Valley jargon, are monetizeable Daily Active Users. They care about how many people they can get to see ads and go to advertisers’ sites.
Since only a small percentage of people who view ads convert and buy something, Twitter needs a pool of many mDAUs as possible, and they need that pool to keep growing so that the pool of potential monetizeable people also grows and grows.
(As a side note, if you ever want to find out what companies care about, look at their 10k statements if they’re public. As someone once said, “When people show you who they are, believe them.”)
The way they market this to advertisers is to put them first:
However, as I wrote in a previous newsletter about YouTube, it’s very much a balancing act between spamming users with Promoted Tweets and letting people create things organically:
So they have to be able to show just enough ads and get people to engage with just enough content to have advertisers be happy, and have the WWIC crowd also be ok about things. Not as much as advertisers, but enough to not leave the platform.
At the same time, Twitter is also competing against Facebook very fiercely for advertising revenue, becoming closer and closer together in the process:
In this context, it’s not surprising that the new site redesign features sections that are clearly paths to get people to click on trending content, hashtags, and other pathways that easier to measure for advertising purposes. The enormous pictures and video expansions that make it impossible to read more than 2-3 tweets at a time are to spur user engagement, which is always higher on visual media than plain text (I’ve found this to be true in my own experience of creating mediocre data science memes as well.)
When someone tells you what they are, believe them, and Twitter is saying over and over again with its redesign that it still doesn’t genuinely understand what makes it successful. Or rather, that it does, but that it can’t make money with that, and is trying to cater to advertisers instead, in a delicate balancing act that it almost always gets wrong.
Art: Dead Birds, Francisco Goya, 1808
What I’m reading lately:
Great post on SQL optimization by Randy:
This:
This is the kind of maternity leave content I need: “Doctors and nurses of Reddit who have delivered babies to mothers who clearly cheated on their husbands, what was that like?”
How to have a lazy summer
More good maternity leave content: I could not stop reading this profile of this Harvard professor who got himself into an ENORMOUS mess
About the Author and Newsletter
I’m a data scientist in Philadelphia. This newsletter is about tech and everything around tech. Most of my free time is spent kid-wrangling, reading, and writing bad tweets. I also have longer opinions on things. Find out more here or follow me on Twitter.
If you like this newsletter, forward it to friends!