Twitter's Fleeting Product Roadmap

This entire post should have been a fleet.

Hi, Can here. Today, we talk about fleets!

I am old enough to remember when Twitter was a tiny company that still put every employee's photo on their employees page. They even had a tiny celebration when they finally had 140 employees. Then, around that time, a Twitter employee got hacked, and some internal strategy documents were leaked to the press. I pored over them, and a distinct phrase stuck in my mind: "Twitter wanted to be the pulse of the planet."

2009!

Like many others at the time, I laughed it off. Back then, Twitter was the website where you'd post about the cheese sandwich you ate. It even encouraged you to post in present continuous tense, so the tweets would read like "Can is eating a salad right now." If you were one of the dorks like I was, those tweets that you'd sync up with your Facebook would make no sense at all. How would a site like that become the pulse of the planet?

Janis Krums @jkrums

5 Year Anniversary of The Miracle on Hudson - Congrats to @Captsully on his miraculous landing. A true hero.

Well, the joke is on all of us. The Hudson River landing maybe was one of the inflection points, as well as its adoption by celebrities (Ellen selfie!). There were rumors about Twitter having to allocate special hardware just for Justin Bieber's account. Twitter would go down so often that its maintenance page with a smiling whale became an internet meme of its own.

It's been years since my last failwhale sighting, which is good. Yet, as I sit here as a 12-year (and probably longer since I had deleted my first account in a fit of misplaced romantic rage in college) user of this service, I can't immediately figure out what has changed about Twitter other than the loss of seeing that ill-fated smile on an error page. What do people at Twitter do?

How I missed thee

I'll be the first to say that at a company of Twitter's size, it can be hard to grasp what all those thousands of engineers and product managers do all day. Often, just keeping the lights on can be a challenge, though one could also argue it should become less so as automation and abstractions take over roles once reserved for humans. Yet, there are a lot of pieces of a product at the scale of Twitter that are just invisible to a casual observer, ranging from law enforcement portals to integrations with various third parties only a few people would have ever heard of.

However, even with that huge caveat, I can't get over feeling unimpressed at every Twitter product launch. I apologize to every single person who's ever worked at Twitter reading this. I am sure it wasn't your fault. But maybe it was?

Twitter's latest launch of its version of Stories, or Fleets as they are charmingly called, is a good case study. As a fan of all things ephemeral, you'd think I'd welcome it with open arms. I think tech companies should store less data as data is as much of a liability as an asset. In turn, people should also actively delete their tweets primarily because you don't want your stream of consciousness to be out in the open forever.

Of course, Stories are a far cry from either. Instagram, which shamelessly copied the format from Snapchat but also ruthlessly iterated on it makes no claims about Story posts being really ephemeral to anyone but its viewers. There's even a way for you to back and see your past stories. Nevertheless, you'd think introducing a native ephemeral format would be a much better alternative than using a third-party tool to constantly delete your tweets as I do.

Twitter Support @TwitterSupport

First, we heard concerns about expired Fleets still being viewable. Fleets are *not* viewable in Twitter apps after 24h. However, our backend system has a queue that deletes Fleets media after 24h –– this system fell behind on Friday morning PST due to scaling problems.

Yet, the entire Fleet product feels so half-baked, at least compared to its competitors, I am left wondering once again what is happening at Twitter HQ (or whatever you call one these days where there are no HQs...collective hivemind? flock?). There were scaling issues on launch, which you can view as a good thing if you are generous but also points to a failure in prediction and inability to respond in time to usage spikes. Somewhat more embarrassingly, especially given the nature of the product, there were even ways to collect all fleets by an account and do it without being even noticed. Using publicly available (though somewhat obscured) APIs, you could turn the entire product on its head.

Joe Weisenthal @TheStalwart

I'm a fleets bull, but product wise feels like a rush job. Not smooth. Janky. I'd assume they will get better, but this is Twitter, so honestly whotf knows. This might be their final state.

But even if you write all that off to usual launch troubles, of which I've had my fair share, the product itself still feels unfinished, if not straight up janky. You can hardly do any of the things you can do on Instagram stories, the interaction to see someone's stories from their account is laughably complicated. There are often loading issues. There are dropped frames during animations, and flickering screens, which all add up that feeling of jankiness. And most concerning of all, at least on my feed, I see very few civilians, those whose relationship with technology is at healthy levels, using it. The entire thing feels destined to be abandoned in a few months.

You could probably write an entire book on missed chances by Twitter and Twitter's main app remains a graveyard of lost chances. It supports not just watching live TV from it, but also posting live videos. It has a full-fledged music player that appears only for certain links, such as SoundCloud (in which Twitter invested with its venture arm). There is a very prominent audio tweet button, which was supposed to be a big thing, yet it's been probably months since I've seen a single one posted. There are probably more things it can do than I know.

I don't know if fleets will follow the same trajectory, but it seems possible if not likely. In essence, the feature feels built not for the users of the platform but for the advertisers who are already marketing other Story-formats. Part of the disjointedness comes from the medium. Stories feel native to Instagram and Snapchat because they are both visual products, whereas Twitter mainly operates on text. Fleets try to bridge that gap by allowing people to share a tweet in a fleet (I hated typing this sentence) and has some rudimentary text tools for fleets too but they are extremely limited compared to those you find on Instagram and the like. The entire product feels unfinished and also stands out like a sore thumb.

What companies like Facebook and YouTube long discovered is that the way to make money off of social media is not to focus so much on advertisers but instead build compelling products that get people addicted to using them and only then sliding in a few ads here and there. You need the users, with their fidgeting eyeballs and gyrating thumbs, before you can get the ads. Twitter simply seems to have built what an advertiser sees on Instagram, not what a consumer uses.

There aren't many companies that have such a crazy ratio of mindshare over iteration. The company is undeniably sitting on some sort of goldmine but cannot even start to find the shovels. For a small, at least compared to Facebook, but extremely influential slice of the universe, Twitter is the pulse of the planet. It is where most public diplomacy happens while also being primary networking tool the tech cabal. It is where people turn to for breaking news and also expert commentary. It’s clear that Twitter doesn’t do catching-up very well. But it can capitalize on what it makes it unique. That would be the real product innovation.

---

Ranjan here: Can and I were discussing what could be an alternative business model that better monetized the sheer level of global attention capture Twitter holds. I still think more commercialized sentiment and data products for news, marketing and financial services, while maybe less scalable, could’ve been far more lucrative than trying to compete with Facebook on straight social ads.