Peter M Howard ::

wintermute.com.au

On Chatter and What Matters

30March2011 /01April2011 [myth]
[tweet this]

In which I wish to discover the information that matters most to me

Some time back, I read an interesting writeup of a post by Twitter’s Evan Williams.

This last point [re. unstructured data] is not obvious but is particularly important for fulfilling Twitter’s goal of helping you discover the information that matters most to you as quickly as possible.

I think this is the first time I’ve seen Twitter’s goal stated that succinctly.

Ev goes on to define the perfect Twitter; emphasis mine this time:

The perfect Twitter would show you only the stuff you care about — relevant, timely, local, funny, whatever you’re most interested in — even if you don’t follow the person who wrote it. And, of course, it would give you ultimate, fine-grained control in how to do so. We want to give you more ways to help the good stuff bubble to the top.

This stuck with me. Social networks have a big problem that I’ve been thinking about for a long time, that of attention, and of value. As an example, consider the relationships I’ve recorded in Facebook. I’m “friends” with family, with people from my past (school, uni, work), and with people I’d call friends today, in-so-much as we regularly see each other in the real world. All of those relationships are at different locations on a spectrum of how much I care, of how much I want to know about the other person. Chatter is important for certain relationships, but for others, there’s little value in the regular noise. In the case of some of my Facebook friends, for example, I’d want to hear about engagements and weddings and children, but I’ve very little interest in their day-to-day lives. In the case of others, I want to hear every little detail. Then most sit somewhere in between.

People in different circles

I’ve few tools available to me, however, to try and manage the level of noise. At its simplest, I just remove people from my Facebook stream (the landing page’s “latest news”) when I don’t want every little detail. This does mean I need to seek out information for other updates though, and that’s not something I want or care to do. Over on Twitter, I restrict the number of people I’m following, and I remove pretty much anyone with only a couple of strikes.

This behaviour serves to make sure I’m mostly only seeing things I care about, but it doesn’t let me see everything I care about, and that’s a crucial gap.

I don’t know what the solution is. Facebook are evidently trying lots of little experiments to improve things in this area, but I wonder if people will tire of the noise before they’re given a useful way to filter things (and “Top News” frustrates because it either filters too much or misses good stuff seemingly arbitrarily). Twitter are working on the issue from slightly different angles. Their location-based Trends are one way, even if they’re flawed. Last Saturday night provided a good example of them working, the trending topics either #earthhour or #nswvotes and other election-related topics. But at other times they’re inane, and they often fill up quickly with spam. Below, election night is on the left, ordinary mid-week on the right.

Twitter trends, election night and mid-week

A few weeks ago, as I returned more fully to Twitter after a two-month hiatus, I upgraded the Twitter app on my iPhone. With the upgrade, I found an annoying semi-transparent bar blocking the view of my timeline with such inane topics. Twitter quickly improved the situation by making the bar appear in a fixed location above the timeline, rather than blocking it, but the update to the app was buggy, with regular crashes as the app refreshed the topics but failed to load my timeline. This apparent preference for the irrelevant was offensive enough to make me find another Twitter client1.

The internet quickly dubbed the bar the #dickbar, and protested2. The protest has been more than just the usual complaints about change, with some great analysis from Marco Arment and Justin Williams. Justin proposes an alternative to the seemingly irrelevant trending topics, referencing Apple’s Genius recommendations in iTunes:

If Twitter wants to improve the trending experience for everyone, it should change how they are calculated entirely. Rather than generating trends based on the location of a tweet, they should instead show trends related to what is happening in my timeline, who I am following and who my followers follow.

At first glance, this seems like a good idea. Justin gives his own example, following Mac developers and mobile computing commentators, so expecting such a solution would show him tech and mobile trends. Alternatively:

If someone follows a variety of sports stars and celebrities, I would expect they see trends related to big games, Charlie Sheen’s latest exploits or Brett Favre’s dong.

Unfortunately, it’s just not this easy. Many of us don’t follow such a narrow circle of people that they would all also be following the same people, which would be necessary to keep the trending topics relevant. And if we did, of course, the trending topics would probably tell us nothing we don’t already know, just acting as an echo chamber. We come back to the same problem I have on Facebook — I follow people on Twitter from all sorts of different circles too, some tech, some politics, some regular real-world friends, with a huge range of interests themselves. Many of the people I follow are themselves following celebrities and sports stars, so their own interests would easily “pollute” my own in any “Genius” trending topics. Even the political tragics are scattered across the political spectrum, and I can’t imagine what I might find trending in those dark niches.

I suspect that, of course, the solution lies somewhere between these. Sometimes it’s going to be useful to see what people the world over are talking about en masse, and Twitter’s Global trends are more interesting to me than their Sydney topics; other times I can count on the people I’m following to re-tweet interesting things, and those act themselves as a sort of “Genius” recommendation, in that I’m predisposed to trust their judgement; other times it’d be useful to get an at-a-glance view of what the most popular topics are being discussed amongst my immediate circles.

Alternatively, we’d have a solution like Fever, which presents popular stories and topics given a whole bunch of RSS feeds. If I could do something similar with Twitter or Facebook, I should be able to friend or follow a whole bunch of people but keep them “muted”, using their streams only to feed a recommendations engine. That might be a useful way to detect trending topics or major events within my wider circles, by looking for the nodes in the conversations.

Footnotes

  1. The hunt for a third-party Twitter client ironically came shortly after reading about recent comments by Twitter themselves that many read as disparaging third-party clients; GigaOm ran Twitter to Client Developers: We’ll Take it From Here; the idea that theirs is the One True Way is especially ridiculous when many of Twitter’s current conventions came from third-party developers; cf Craig Hockenberry’s Twitterific firsts  

  2. Addendum 1/April: Just this morning, Twitter removed the QuickBar completely from their iPhone app; their blog post announcing it is worded such that it could be responding to any unpopular change though, and misses the point somewhat, talking up the concept but failing notice the weakness in the execution  

« 'Sucker Punch' and the Decline of Strong Woman Action Heroines :: Le Flâneur »

Related [myth]