Twitter_512x512Twitter yesterday published this post about their continued “experimentation” with people’s Timelines. Specifically they are looking to keep finding ways to bring in tweets that you “might miss out on” but which may be interesting to you. And here’s how, according to the post, they’re going to make that determination:

  • Activity on accounts you already follow
  • Popularity of the Tweets
  • How people are interacting with those Tweets

But then it ends with this thought:

“As the timeline evolves, we will continue to show you Tweets you care about when they matter most.”

I completely understand what Twitter is trying to do. They want to make the experience, specifically the on-domain experience, more engaging for new and non-power users by bringing up material they feel is relevant. So they’re using those signals (and whatever else) to judge what people are mostly likely to find interesting and, as they say, fill holes in the Timeline with those posts.

But the problem is that the final statement quoted above doesn’t jive with the idea of an algorithmic feed. By definition an algorithmic feed shows the posts *other people* care about, making the assumption that by association you will as well. If other people find it interesting you will too, so we’re showing it to you is the logic.

Let’s start with a simple premise: Social networks, when you first set up a profile, are an opt-in mechanism. If I want to get updates from X (where X can be either an individual, a brand or a fictional character) then I will take the positive action of clicking the Follow/Like/Subscribe button. These are the updates I want to see because, for whatever reason, they’re important or interesting to me.

Then one of two things happen: Either users keep following more and more accounts until their feed gets overwhelmed or they decide that, nope, that’s good and stop at just a select few. (Note that I’m excluding those who completely give up, either deleting their account or abandoning it and becoming “Inactive” accounts.)

Those in the second group may not feel any need to see more posts. They’re good where they are and anything additional is likely to be confusing and unwelcome. For those who fall into first group, tools already exist to help them determine a good signal-to-noise ratio, most notably Lists. But Lists are something that’s not a great user experience either on-domain or through Twitter’s native mobile/desktop apps. The best option there is Tweetdeck or an outside tool like Hootsuite. There are options though.

Plenty of us like the stream, even if we also use tools that allow us to curate based on our interests. The key, though, is that those tools are in our control. If I want to remove someone from a List I can. I don’t need – nor do I want – Twitter determining who it feels I should see as important. Just because something is popular among the people I follow doesn’t mean I need to see it. NYU’s Jay Rosen has dubbed it the “Ice Bucket Feed” since it’s more likely to show posts that have been engaged with by others, as the Ice Bucket Challenge was on Facebook, but engagement doesn’t automatically translate into importance.

To be clear, I’m not against Twitter trying to increase engagement and subsequent usage. But, as I’ve stated before, it needs to at the very least have an opt-out option that’s permanent until I say otherwise. One of the constant frustrations is every day going to Facebook and LinkedIn and changing them to “most recent” as opposed to the feed that shows what it thinks I’d be most interested in. I don’t want to outsource what I see to these networks.