Ever since Twitter unveiled its new Fleets, I’ve been mulling over the company’s future. What can they do that’s uniquely their own? How can they take advantage of their past and present in a way that builds a better future? Needless to say, Fleets — a watered-down version of a content-sharing format introduced by Snap and already copied by Instagram, Facebook, and others — is not the answer.
Instead of being the latest in a growing line of companies that have pinned their hopes on Stories, Twitter could improve their prospects by exploring another model —one that could also help them lead the way in improving the entire media ecosystem.
But before delving into how they should move forward, let’s look back at some of Twitter’s history, which just happens to intersect with my own.
***
In July 2006, I attended a party at Ruby Red Labs in the SOMA district of San Francisco, where one of Twitter’s founders gave me a demo of the product. The tweet I sent that night was the first one ever sent by a non-Twitter employee. (I wrote a piece about the new service on my old blog later that night in a mildly inebriated state, which is apparently not when I do my best spelling.) The earliest version of Twitter did only one thing: It broadcasted your status to your friends via text message. My accidental launch of Twitter gummed up my phone with so many text messages that I got in trouble with AT&T.
Twitter’s vibe was very casual at the time, and it remained so up until the service blew up at South By Southwest, the annual music and technology festival down in Austin, Texas, in 2007. If you had many friends at SXSW that year — and I did — they sent a lot of text messages, and Twitter quickly became a very noisy nuisance. One of my friends in the mix (and an unpaid Twitter R&D guy), Narendra Rocherolle, decided to build a WAP (back in those innocent days, this simply stood for Wireless Application Protocol) version of a personal feed.
On the desktop web, early Twitter had a homepage that served up a log of every update on the system. In fact, as part of their marketing at SXSW that first year, they put up plasma screens showing off the global feed of updates. For a while, the company allowed experimentation around desktop clients and looked for a better way to display this stream. Does anyone remember Tweetie?
But it was never clear what product Twitter wanted to build. They had an API and client ecosystem, but they had no clarity around what they wanted to be. Ultimately, they opted for the media model and began selling advertising against engagement.
While this shift may have helped the company get ready for the public markets. When they went public, I wrote that “Twitter is about to enter that weird zone where the need for growth trumps everything, including one’s values. That is Twitter’s fork in the road.”
It put them on the unfortunate path to what they have become — a platform whose core principles of amplification and engagement are suitable for news, disinformation, marketing, and disjointed conversations. Like Facebook, its more successful and more evil older brother, Twitter is built around dopamine loops. It encourages behaviors that are mostly mindless and driven by emotional triggers.
Given its roots as a loosely defined conversational social network, Twitter is neither good at being a conversation platform nor useful as a distribution and discovery platform. Trying to conduct a conversation on Twitter is like being trapped inside a Tower of Babel while encased in bubble wrap of self-promotion.
Be it good, bad, or ugly, distributing content on Twitter is easy. Meanwhile, the discovery has been alarmingly bad and getting worse. No matter how good its intentions, nothing in the company’s core product and behaviors reward quality over quantity in its rapid loops of information.
But it doesn’t have to be that way.
If Twitter is willing to rethink its entire core application, jettison the past, and build a new value system around content, it could create a more relevant, robust, and financially rewarding future. And while it shouldn’t look to Facebook, Instagram, and Snap for inspiration, it doesn’t have to look very far.
Twitter should be setting its sights on Spotify.
BE THE SPOT LITTLE BLUEBIRD
What is Spotify? When you abstract the service to its basics, it is a content distribution platform with a payment system. Spotify’s database allows it to tag, sort, and share content — specifically, music and podcasts — across three axes:
- Artist
- Playlist
- Genre
It is a database that tracks who listens to whom and, subsequently, who gets paid how much. It creates demand and decides who gets the rewards based on listening behavior. The revenue to share comes from either subscriptions or advertising. In recent years, they have started to make more money from subscriptions, so their revenues are beginning to become more predictable.
Spotify lacks a robust social network, which limits its ability to turbocharge the distribution of this content. Instead, it relies on discovering content differently — either through individuals and their playlists, artists and their playlists, or auto-generated radio stations. Many of Spotify playlists are controlled by the company, and like radio stations of yore, they have thus become kingmakers.
One can’t deny that Spotify’s platform is an uneven playing field. It skews in favor of the biggest, bold-faced names. Frankly, it doesn’t work for most artists. However, one also can’t deny that it has been a key reason consumers don’t think twice about paying for a music subscription. If not for Spotify, Apple would still be selling us downloads.
Twitter, on the other hand, is a distribution platform for multiple content types: Written words, audio streams, and video streams. It can generate attention for a news article, a podcast, or a video clip. It doesn’t differentiate between different content types. It can’t tell whether something is real, marketing, or just completely fake. In recent times, this content discovery has taken a hit, thanks to bots, spam, and trolls. Unlike Spotify, it lacks a payment system, so it has no mechanism to enrich those who participate in its ecosystem. It relies on one single source of income — advertising, whose growth is showing signs of fatigue as monetizable daily usage is starting to plateau.
In July 2020, Twitter CEO Jack Dorsey mentioned that it was looking at offering a subscription model. Given that Twitter stock ticked up on the news, it tells me that Wall Street approves of this non-advertising revenue stream. The company has started thinking about features that could be paid options, but most of the proposals make me cringe.
Rather than proceed along these lines, they should take some lessons from Spotify. Twitter should develop a second revenue model that has the same potential as its ad business. This revenue model should be focused on the centrality of news, information, and content to the Twitter experience. In addition to allowing Twitter to grow, it must allow an ecosystem to thrive.
Twitter has one advantage that it needs to press: Many experts, all media companies, and most significant sources of information already exist on its platform and use it to disseminate information. Many Twitter users also know that news happens on Twitter. To push the Spotify analogy, all the artists, record labels, and their fans already live on Twitter. This is a unique advantage for the company.
We need an elegant discovery platform for news and information, and Twitter could be the company to provide it. Doing so involves deprecating the classic all-updates feed and pushing it aside. This would provide space for a new platform that — like Spotify — allows quality and relevant information to bubble up. Instead of artists, playlist, and genres, Twitter’s content could be organized around these three axes:
- Brands (Artists)
- Topics (Playlists)
- Genre
For example, The Washington Post is a brand. It can leverage its various content types to provide links to its articles, tweets from its editorial team, and even live discussions. Some of this content — coverage of a breaking news story, for example — might show up in the streams for specific topics, which function somewhat like playlists.
An individual, say an independent cricket journalist Melinda Farrell could be a brand, create her own content and playlists. A company like American Express, for example, could create its own content experience and replicate all its offline efforts.
An example of a genre might be “Yankees Baseball.” Twitter could aggregate the best accounts to follow and the latest articles, and this would help users discover good content. Twitter already offers basic and frankly very poor versions of this offering. I use many of them and find them more useful than the main feed.
Twitter, actually is doing well with Topics — or at least automatically generated topics. In its most recent quarterly call, Twitter pointed out that “The number of accounts following Topics also grew significantly, reaching 70 million by the end of Q3, up 40% quarter over quarter.” This is a very strong indication that this behavior works. If they can give humans the ability to create their own unique topics (aka playlists), they can only make the experience better.
Twitter, like Spotify, should focus on a highly effective recommendation system, that leverages the human input on the platform for discovery and recommendations instead of wasting its resources on me-too features.
In this imagined future-Twitter, the interface would only surface content from areas of the platform of which users are a member or a follower. The owners of brands, topics, and genres could either get a percentage of ad revenue or charge subscriptions (with Twitter getting a fee, of course) for access to certain types and pieces of content.
In much the same way Spotify has become a place where people experience music, Twitter could be the place where we discover, share, and consume news and other written content. And unlike Spotify, it could be a place where new, independent voices are found and build an audience.
Spotify’s reward structure doesn’t help the independents, and many smaller artists feel left out in the cold and understandably frustrated. Twitter could develop a subscription system that rewards both big and independent content creators. A system proposed earlier could be a skeletal template that satisfies both the big and the independent.
There were rumors that Twitter was looking to buy Substack, but the company shot that rumor down. If I were in Jack’s shoes, I would think twice about that. There is plenty to learn from Substack. It would be prudent for Twitter to see which media voices have an engaged audience on their service and help those people make money off that audience.
So, what’s missing? Given that it already owns news discovery and the social conversation, future-Twitter needs are a payment layer that would allow money to connect to the media ecosystem. For news consumers, a monthly subscription that unlocks all the news would solve the paywall problem.
At this point, I am feeling subscription fatigue and have no desire to sign up for any more annual subscriptions. It is like buying twenty-five albums to listen to twenty-five songs. Instead, I want an experience like Spotify, where my subscription pays for all the music I enjoy. I want my Twitter subscription to pay for what I am reading. By the way, this would be a much better experience than other half-brained subscription products, like Apple News.
Twitter needs to change gears quickly — and it needs to start rebuilding itself now. It won’t be long before the toxicity on the platform starts to deprecate the brand itself. So far, the company seems to be set on taking the content-delivery road more traveled. Changing course could make all the difference.