There was a lot of attention on Twitter on Monday, as the company reported its quarterly financial results: despite meeting Wall Street estimates on things like revenue, the network’s share price slid by almost 10 percent. According to most of the stock experts who commented on the slump, investors were looking for better user growth and engagement numbers — and that dilemma helps explain why Twitter is going to continue to muck around with your timeline using algorithms, regardless of whether users want it to do so.
The first hint that this trend was likely to accelerate came during an interview last month with new chief financial officer Anthony Noto, who said that the traditional reverse-chronological timeline arrangement “isn’t the most relevant for the user.” Although my summary of his thoughts drew plenty of fire from Twitter and its defenders — who argued that I had taken his comments out of context — it seemed clear to me what was coming.
— dick costolo (@dickc) September 4, 2014
In fact, comments made by both CEO Dick Costolo and by Anthony Noto during Monday’s conference call with financial analysts made it clear that these experiments are going to continue, and if anything will be broadened to the point where they will likely impact every user to some extent. If you see tweets that have been favorited or rewteeted by people you don’t follow showing up in your stream out of sync with your timeline, then you will know that you are part of the rollout.
Can Twitter have its cake and eat it too?
The company is doing its best to make it sound like this is not a big deal, and that the reverse-chronological timeline will remain intact. But the reality is that for many users, the ability to curate their own experience on Twitter is a crucial feature — just take a look at the survey we did in September, which was 87-percent negative on the idea of an algorithmically-curated feed. The idea of someone else deciding what’s important in their stream appears to be anathema to many users.
Of course, for Twitter this probably feels like just another enhancement that they think will make the feed better for some — primarily new users and those who don’t log in very much, which is a key market for the kind of future growth Wall Street wants to see. And the stream has already been disturbed by things like promoted tweets and other forms of advertising, which show up out of sync with the timeline. What’s the big deal about one more disruption?
Supporters of Twitter’s move, like early investor Chris Sacca of Lower Case Capital, argue that the changes will make the service more user-friendly, especially for new users, and that existing or power users shouldn’t see it as a threat. But it’s become very clear that some don’t want Twitter to alter their stream via algorithm even if it claims to be doing so for their own good.
The slippery slope of algorithmic filters
The most common response to the idea is that it will make Twitter like Facebook, where an algorithm routinely promotes or down-ranks content based on criteria that users can barely understand — if they even realize that their feed is curated at all (which surveys have shown that many do not). In an impassioned plea to Twitter not to implement a Facebook-style algorithm, sociologist Zeynep Tufekci recently summed up the feelings of many die-hard Twitter users:
[blockquote person=”” attribution=””]The key to this power isn’t the reverse chronology but rather the fact that the network allows humans to exercise free judgment on the worth of content, without strong algorithmic biases. That cumulative, networked freedom is what extends the range of what Twitter can value and surface, and provides some of the best experiences of Twitter.[/blockquote]
Although Facebook’s feed is different in that it actually hides or removes content without telling people, based on whatever signals its algorithm chooses to look at — few of which are publicly disclosed, of course — a number of the company’s fans have pointed out that Twitter is only talking about adding content to your stream via algorithm, such as “important” tweets that you might have missed.
But as more than one person has noted, this is a slippery slope at best: once you have made the decision to alter the flow or the arrangement of tweets based on abstract criteria that have been programmed into an algorithm, then at some point you are going to decide — for the user’s own good — to hide or remove certain things as well, because you want to improve the signal-to-noise ratio.
— Ravi Somaiya (@ravisomaiya) October 27, 2014
The ironic thing about Twitter is that even though many (including me) complain about the difficulty of plowing through all those tweets and finding the signal, we very much want to be the ones doing the filtering, rather than having it done by a faceless algorithm. The whole point of having a social network in the first place is that the people you choose to follow are the algorithm. Flawed, perhaps, but human, and therefore somehow wonderfully unpredictable.
Encouraging more engagement from new users makes perfect sense for Twitter as a company, and particularly a public one, which has to justify its market value to investors. But the changes that are required in order to do that may not make sense for many of the network’s long-time or power users — and that’s a dilemma the company is going to have to confront sooner rather than later.
Post photo and thumbnail courtesy of Thinkstock / Ociacia