Why We Should Embrace Twitter’s Experiments Instead Of Shooting Them Down

Edge’s Richard Parker says Twitter is going to change the way it delivers its timeline whether users like it or not so we may as well concede the tech platform has a point.

Apparently, Twitter’s chief financial officer (CFO) Anthony Noto, has been running his mouth off at the Citi Global Technology Conference in New York, about Twitter’s imminent timeline rejig. He was quoted as saying, somewhat undramatically, that the reverse chronological order of content that Twitter is known for “isn’t the most relevant experience for a user”.

Apparently Twitter tested the idea of surfacing content that people you follow have favourited in your timeline back in August – and it didn’t go down to well amongst Twitter users who didn’t take kindly to what they saw as Twitter forcing content on them that they hadn’t asked for.

So what exactly did Noto say? Well, according to the Washington Post:

It’s important to understand what Noto did and didn’t say. He pointed to the problem of important content disappearing quickly and said that Twitter might resurface tweets from people who the user already follows if the company deems those tweets important. How that would work is anybody’s guess. But Noto did not suggest that Twitter would start filtering or curating timelines (my emphasis). If all Twitter does is resurface a tweet from someone that a user follows, they’re not distorting what the user sees. That user already follows the person whose tweet they’re now seeing. They would have already seen the tweet if not for the fact that they were away from their timeline.

Hmm. Noto’s clearly got a point about content disappearing quickly on the newsfeed. It’s problematic in a number of ways.

Firstly, there’s a threshold – a function of the number of people a user follows and the frequency at which those users post – after which without being logged in all the time, users will miss a lot of posts. The user has a choice here – accept missing the posts, or follow less users.

And obviously, from a broadcast viewpoint – whether you’re a brand or an individual – it’s difficult to get content out to your entire follower base without tweeting the same content numerous times, which could alienate users who spend a lot of time on Facebook.

These are issues. But is resurfacing favourites-of-friends the answer? I’m not so sure about that.

For one thing, just because I follow someone doesn’t mean that I care what they favourite. I might, for example, follow a politician for their views on government – I don’t care that they favourite a brand or a musician or a place, and yet if they do, the content from that user will be surfaced to me. No thanks. It also affects my behaviour as a user. Am I more, or less likely to favourite something knowing that it might then show up in my friends’ feeds?

For another thing, not all of the people I follow are equal in terms of their influence on me – or to put it more succinctly, I give more of a shit about some users’ opinions than others. Twitter would need to do some serious work to make sure that they only resurface content from favourites of friends whose opinions I respect. And even then, they can only go on observed behaviour, and what I click on or favourite depends on what I see on my timeline at any given moment – which is a random, time-based sample of the users I follow, and not necessarily reflective of my rational preferences.

I wonder whether Twitter could take this opportunity to do something more interesting. It would be a shame to see them reduce the amount of content that users see, as Facebook has done, based on a bunch of assumptions. Rather than ape the algorithmic tinkering of Facebook, Twitter could allow their medium to become truly user-controlled by allowing people to prioritise the users they follow – a simple preferential indicator like ’ Trusted’ or ‘Respected’. Giving users this power would also be interesting from a brand viewpoint: only brands that consumers really want to hear from would get any kind of cut through.

This approach COULD help to negate the filter bubble effect, too. Sometimes our actual behaviour is different to the way we think we want to behave (that old thinking fast and slow thing), so an algorithm that shows us content based on observed behaviour could be reducing the breadth of opinions that our evolved brains want to see. But it would also reduce Twitter’s opportunities to eventually ape Facebook in gradually reducing organic exposure for brands, and helping to position Twitter as a paid platform.

Whatever happens, something is sure to change, whether users want it or not – and experience tells us that any tinkering with timelines and news feeds is greeted with initial uproar, gradual acceptance, and eventual amnesia that it was ever different.

Richard Parker is the joint managing partner of Edge.