Web dev at the end of the world, from Hveragerði, Iceland

Why I am worried about Twitter and why you should be too

It’s very easy to look at an organisation from the outside—to look at its position, strength and weaknesses—and come up with strategies that build on those attributes.

It’s so easy that even Hacker News commenters can do it.

If Company X started bundling up services from its partners with its own offerings that would give them an unassailable position in the market and they’d make a lot of money!

If Software Co Y would just add this feature they’d be minting it!

If Platform Z just opened up they’d massively increase their revenue!

They’d all be true statements and then the commenters get into an argument about which strategy is the ‘correct’ one. They’d flame each other, not realising that strategy that’s been suggested in the thread would probably work for the company in question.

Provided that:

  1. The company actually picks a strategy and acts on it.
  2. And that the company is actually able to execute it.

Plausible strategies are cheap and abundant. It’s the primary output of the various empty-suit MBA graduates that litter the modern corporate landscape. Some strategies might be more optimal than others, but even a second rate student from a third rate business school will come up with a strategy that’s in the same ballpark as the optimal one. The modern business world doesn’t lack strategies.

Inability to execute the strategy that the management has chosen is what kills companies. Most of the time it’s because a competitor or disruptor has pushed the company into a position where it has to work outside of its competencies. But sometimes it’s because the company in question is a startup and it just hasn’t managed to build up a fully functional organisation in the first place.

Twitter’s investors clearly believe that its problem is the former, that it’s a company that has been pushed out of its comfort zone by a stronger competitor: Facebook.

My worry is that it’s the latter: Twitter simply hasn’t managed to build an organisation capable of executing the strategies necessary to build and maintain a successful social media platform.

Leaving aside the actual strategy—I have my opinions on what I think Twitter should do but that’s irrelevant at the moment—there are four reasons why I’m sceptical about Twitter’s ability to execute any of the strategies people have been proposing.

1. Dysfunctional software development

If you look at Facebook’s software development efforts, they keep shipping one paradigm-shifting idea after another.

Facebook’s capability for software development is easily on par with Google’s or Apple’s.

Twitter, in comparison, isn’t even able to keep its own platform up to date. The Mac OS and iPad apps are an embarrassment. Even on iPhones, Twitter is regularly outshined. Tweetbot and Twitterific habitually ship apps that are better designed, less buggy, faster, more responsive, and better integrated with iOS in general, than Twitter’s own app. This is despite these third parties being deliberately crippled by Twitter.

Twitter’s inability to ship native apps would be tolerable if it were able to keep its web app on the cutting edge, but even there it is incapable of shipping important strategic features that it’s betting the company on (i.e. Moments).

Twitter has a demonstrated inability to iterate its own software platform. Even a small thing like its supposedly temporary ban on URLs in direct messages ended up lasting more than a year. When a temporary fix lasts for months, you know there is something broken in the development process.

Facebook is an unusual company. It’s big, has excellent development teams, and seems to know what it’s doing. Three things that rarely seem to coincide in one organisation.

Twitter is setting itself up as a direct feature-by-feature competitor to Facebook with its switching from favs to likes and a plan for an algorithmic timeline. But it hasn’t demonstrated even a shred of the capability necessary to stand toe-to-toe with Facebook. It’s entering a boxing tournament and signing itself up in the wrong weight class.

2. Inability to tackle harassment and abuse

If you don’t think harassment is a problem on Twitter, then you probably don’t follow that many women or members of minorities. It’s a huge problem.

  • Twitter’s reporting tools are too limited for a problem this complex.
  • It’s slow to respond to reports.
  • It’s inconsistent in how it handles abusive accounts.
  • Users have limited tools for dealing with abuse themselves (i.e. limited blocking tools, inability to consistently block media and images).
  • The blocking that it offers doesn’t work (even when logged in, the blocked account can still easily find the blocker’s tweets).
  • Its policies are vague, only cover a subset of abusive behaviour, and frequently comes across as condescending towards the abused.

Abusive users often lose interest once they realize that you will not respond.

Gee, thanks. I’ll try that next time!

People who are targeted by abuse and harassment (i.e. women and minorities) are currently staying on Twitter purely on the promise that this will be addressed. The longer that Twitter lets this wound fester, the more likely it is to start seeing a substantial loss of active users.

Which would kill off its advertising business because no advertiser is interested in a platform that only has straight, white, middle class, male geeks.

(I cannot over-emphasise how much of a ticking time bomb Twitter’s harassment problem is.)

3. Inability to build a good user experience

The wonderful thing about tweets is how easily sharable they are. You can embed them in posts. You can throw a link to them in a chat with your friends or the cat gif channel in your workplace’s Slack. A tweet is a short very sharable comment that sometimes links to something interesting. It’s what sets Twitter apart from other social media. The other social media sites tend to be more like silos.

Moments break all of that by simply not working on the web and by hiding all of the links that each moment aggregates. This cuts off both legs off Twitter’s own core competencies:

  1. Users don’t share Moments because a lot of the people they send tweets to can’t see them.
  2. Publishers don’t share Moments because they hide the links to their articles.

It’s an interesting idea that never should have been shipped as it’s currently designed. Simply put: it’s a bad user experience. A Moment should have been an easy way to keep up with an ongoing event, but the link hiding makes it useless for that purpose. Moments seem designed to out-viral even the most viral Buzzfeed article but never even get started because they don’t work on the web. It is a great idea with dumb execution.

This inability to design and execute is pervasive throughout Twitter:

  • Polls are invisible in clients that don’t support them (not even a link). Which wouldn’t be a problem if they were able to keep their own apps up to date (see above).
  • And polls are a useless gimmick as shipped given that they only allow for two possible answers.
  • It’s been years and @mentions still occupy character space in the tweet itself, completely undermining the social aspect of the platform. Conversations quickly become very limited as the mentions start to fill up the tweet itself. No matter how long you think tweets should be, this is an improvement that has been obvious and necessary from the start.
  • Same with the fact that links/URLs still take up character space.

What’s worse is that the features it is chasing show a distinct lack of vision. At a time when Facebook itself is reevaluating how it presents and handles its like button and when Slack is experimenting with new approaches like emoji reactions Twitter switching to likes/hearts is a case of skating to where the puck was, not where the puck is going.

Given how badly Twitter has been designing their recently shipped features, I have no faith in their ability to ship something as important and central to its success as an algorithmic timeline. Its investors should be worried as well.

4. Layoffs

Sacking a large part of your organisation is always disruptive. Always. It doesn’t matter if you do it well (which Twitter didn’t). It doesn’t matter if it was necessary. Layoffs always disrupt existing projects and plans. The only way to minimise the risk inherent in layoffs is to cancel a boatload of projects at the same time. The company after layoffs should be doing a lot less than before. That’s the point of layoffs.

And no, you can’t just ‘trim the fat’. Everybody is working on something in a company of Twitter’s size. They may not be productive but they’re all doing something. If you’re laying them off and not cancelling their projects you’re just increasing your organisation’s dysfunction.

Layoffs have another unique risk for software companies: they increase the odds of a developer team death spiral.

The cycle is incredibly common:

  1. You have a dysfunctional team that has problems shipping features.
  2. Management addresses the problem by announcing a new strategy and appeases investors by introducing layoffs.
  3. The development team now has an increased workload (i.e. new strategy while winding down the old one), fewer developers to handle it, and less support staff.
  4. Development continues to be dysfunctional, leading management to push for death march ‘heroic effort’ tactics, which lowers moral and increases the dysfunction.
  5. Developers quit because moral is low and the inability to ship has pushed the value of their stock options down. Recruitment becomes a problem for the same reasons.
  6. Management responds to more dysfunction by pushing the team harder, further increasing the dysfunction and further lowering moral.

And it just cycles on from there.

It always starts with well meaning ‘short term’ efforts.

We just need to push a little bit harder on this project. We really need this to ship. Things will turn back to normal after we ship.

But it never does.

Before you know it the death march becomes the norm and nothing ever ships until the company gets acquired or goes bust.

Twitter’s layoffs may well have been necessary, but doing them now instead of after addressing other core problems (or cancelling a boatload of projects and apps) is a huge risk to take. It’s a bet-the-company gambit that they took just to appease its investors in the short term.

(How do I know that Twitter hasn’t been cancelling projects? I don’t. But if they have been, they haven’t gone for obvious candidates like end-of-lifing stagnant apps and platform features.)

It doesn’t look good for Twitter

The company is facing a series of huge risks and it hasn’t demonstrated to its users or its investors that it is taking them seriously.

Until it does, it doesn’t really matter what strategy it chooses because the odds of them pulling it off are very slim.

This tweet by Saladin Ahmed sums it up:

My whole professional life (and a lot of my social life) is centered on twitter. Changes that look like they might sink it stress me out.

— Saladin Ahmed (@saladinahmed) November 3, 2015

I don’t want Twitter to fail. I want it to succeed because, like Saladin, it’s where I’ve centred my online professional life.

Maybe they’ll pull it off. Maybe they’ve already got a handle on their dev and design dysfunctions. But nothing they are saying or doing publicly makes me think so.

There is nothing to indicate that they understand the risks they are taking.

And, like Saladin, that stresses me out.

(Have I been using the sports analogies correctly? Hope so. Everything I know about sports comes from people using it in analogies I read.)

You can also find me on Mastodon and Twitter