I recently had a discussion with Zac Cichy on the benefits of third-party Twitter clients. It was enlightening, since it became apparent that both of us use Twitter in very different ways. One reason why I use Tweetbot and Nuzzel for Twitter is because I want two different Twitter experiences: a social experience, and an informational experience—and I want them separated. Zac wants a combined experience that surfaces the best content. I don’t trust Twitter to do that right, and prefer to miss out rather than get irrelevant info. Then again, I don’t trust anyone to get that right.

That there’s no one way to use Twitter is the core of Twitter’s problem, and also the solution. In a world where Facebook dominates social with a one-size-fits-all platform, a flexible and extensible platform is a harder sell to investors. If Twitter decides to embrace it instead of becoming more like Facebook, however, it will completely upend the way we do social media. Imagine, if you will, a Twitter that worked like plumbing to a robust ecosystem of apps that slice and dice your feed into useful information. Twitter could be the back-end to a news aggregation service, a chat room, a way to share media, and more all depending on the apps you hook into it.

If this sounds a bit like the App.Net elevator pitch, you’re right. App.Net, though roundly viewed as a Twitter clone with an entry fee and longer posts, was supposed to be a platform for social apps. This didn’t work, and it was largely because the public image of App.Net ended up myopically tied to the paid Twitter clone part. The rest was because the interesting apps to leverage the network never really materialized. Just look at the App.Net Directory. I recall a Foursquare clone, an Instagram clone, a notes app, a file sharing app, and several Twitter-like clients, but all were hampered by ADN’s poor reach. Fortunately, if there’s one thing Twitter has, it’s reach. Twitter is where stuff happens, where people communicate—even if those people are often celebrities or presidential candidates.

To make this work, Twitter needs to rethink the basic experience of the platform, though it won’t be all that dramatic of a rethink. The current, out-of-the-box, Twitter experience sucks. It sucks for new users, it sucks for power users, and it sucks in a lot of the same ways for both. For new users, Twitter is confusing. Not just in terms of terminology, but in how to get the most value out of it. They’re the ones who would benefit most from a robust app ecosystem. If a new Twitter user wants news, peeks into celebrity lives, or just to gossip with friends. Power users need a more flexible platform and APIs to both make the apps and use their feed in better ways.

Twitter’s first-party web and mobile apps need to become gateways to different ways to get more out of Twitter. They could cram all of these ideas into the main app, in much the same way they’ve crammed Moments in, but that doesn’t solve the problem. Instead, Twitter should focus on four core apps, and let third-party developers do the rest. Those apps are:

Timeline

This is the existing Twitter app: a straight chronological firehose of tweets, possibly with optional algorithmic surfacing of content you missed. The focus of Timeline is on the “core” Twitter functionality of Follow, Post, Share, and Favorite. Timeline would also serve as the gateway to Twitter’s other app offerings, first- and third-party alike. If a user starts following news sites, push Twitter News as a way to get more and better news. Did they import their Facebook friends? Twitter Chat.

Twitter Chat

Why this doesn’t exist yet, I simply cannot imagine. If there is one app Twitter desperately needs to make, it’s a private messaging app. Twitter’s Direct Messaging feature is one of its strongest assets, and they recently rolled out group messaging. A chat app would also give people who are worried about posting publicly a way to use Twitter socially without the risk of being mobbed. Chats could also have the option to be semi-private, allowing for jump ons, while keeping it segregated from the main timeline.

Twitter News

A lot of people use Twitter for news, so Twitter should have a content consumption app with a focus on links and media. (Pro Tip: if a Twitter executive is reading this, buy Nuzzel and you’re halfway there.) Twitter News would both surface links of interest shared in your timeline, by friends of friends (another Nuzzel feature), or from curated lists based on topics. Users can either follow those curated lists, or build their own based on news services they prefer.

Twitter Video

Somewhere between Vine and Periscope, Twitter Video would let you post, share, and stream video. A similar app might be good for photos, but Instagram has that market locked up. I won’t say Twitter shouldn’t try, but I think they have a better shot at Video, building on the success of Vine and Periscope. Twitter Video can also feed into Twitter News to aid in citizen journalism: a user can go from a live, amateur stream of an event, to Twitter News for the official story, and back.

Beyond

Third-party developers could pick up from here, though as long as we’re entertaining fancy, I would love a Twitter Music app. (Yes, they had one before, but it sucked.) A good Twitter Music app would let me follow the bands and musicians I adore, and keep me up to date with upcoming shows, new releases, and general band gossip. You know, like Apple Music Connect, except not terrible.

Another useful idea: Twitter Local—aggregating local Tweets and trends so you can see what’s going on nearby. Something I miss, dearly, from the early days of Twitter, was being able to see who near me was tweeting, especially if something was going on in my neighborhood, like cops swarming an intersection. How about CelebriTwitter: an app just for following celebrities? A Twitter dating app? A Twitter app for finding out what events are going on nearby? With a robust API and enough exposed data—most of which already exists, the possibilities are almost limitless, and it might not even be all that hard to monetize it.

These ideas alone won’t make Twitter not suck. There’s no shortage of low-hanging fruit Twitter can knock off to improve the existing experience. Hell, I’d probably switch from Tweetbot to the official Twitter client in two heartbeats if they just provided a goddamn Edit button for when I make a stupid typo in a Tweet. Twitter needs to embrace the fundamental flexibility of its platform. It’ll be a much harder sell to investors than simply copying Facebook, but the potential payoff could be huge.