We’ve done an awful lot of talking about weibo real-name registration here at Tech in Asia, because it’s both interesting and important. But one thing we haven’t talked about is how Sina’s half-assed approach to the real name rules has affected how many people are actually using the site.
While it’s a far cry from having any official numbers, I’ve been tracking a few key metrics since late February. Now that real name registration, at least in its current iteration, has been in place for a while, I thought it might be useful to share these numbers.
I tracked five very simple statistics: the total number of posts about the top two trending topics on any given day, the posts-per-minute count for those same two trending topics, and for kicks, the posts-per-minute count for my own personal weibo feed. I recorded all five of these stats once each workday at 2:00 P.M., from late February through last week (although I missed a few days here and there when I happened to be away from a computer at 2:00).
Now, before we start, you should take these numbers with several grains of salt, for a bunch of reasons. First of all, I collected all this manually, so there’s a chance of human errors in the data, (as evidenced by the fact that I missed a couple days). Also, as the trending topics aren’t entirely automatic — Sina can and occasionally does artificially insert or delete topics from their list — looking at them might not really be representative of the overall discussion on weibo. But with very limited time, I figured these were good, simple things to track and if there was a huge drop-off in traffic, it ought to be evident even in this flawed data.
I’ve posted charts and my own conclusions below, but you can also click on the charts to view an interactive version or take in the raw data right here. You’ll note that in order to make the charts readable, I’ve had to toss out the numbers from a few days with extremely high post counts. For example, a topic about Tomb Sweeping Day on Tomb Sweeping Day was running at more than sixty posts-per-minute; it does not appear in the graph because including it would have made it difficult to actually see any trends that might exist on more “regular” days when counts are generally much lower.
Also, for reference, the real-name rules began going into sorta-effect on March 18, and seemed to be being rolled out gradually over the next few days, so that’s the point you want to look for to see if there’s a noticeable difference before or after it.
By Total Posts
How much are people saying about the hottest social media issue of the day? Would those numbers decrease after the real name rules went into effect? I measured the total posts for the top two trending topics at 2:00 PM each day, and here’s what I found:
Click to view an interactive version of this chart.
Click to view an interactive version of this chart.
Looking at these two charts, it’s difficult to see much of a pattern. There’s a definite downward slope in the number of total posts about the trending topic number two, but it isn’t mirrored in the results for number one, so it’s impossible to tell which is a more accurate depiction of what’s really happening. From the two of them, though, we can at least tell that there apparently hasn’t been a massive drop in traffic or usage, and the most interesting topics are still attracting huge numbers of posts.
Posts-per-minute
I also tracked posts-per-minute for the number one and number two trending topics using a very simple method: counting. Since all weibo posts have a timestamp, it’s actually pretty easy to figure out how many posts appeared in a trending topic’s feed within one minute’s time. What I generally did was pick a minute around 2:00 PM and count the number of posts made in that minute, and then double-check with a few other minutes around 2:00 just to be sure I hadn’t gotten a fluke and that the number I got was generally representative of how frequently people were posting about the topic. Unscientific, to be sure, but it’s better than nothing, right? Here we go:
Click to view an interactive version of this chart.
Click to view an interactive version of this chart.
Here, too, it’s hard to see a definitive pattern. Maybe there’s a slight downward trend there, but it’s definitely nothing earth-shaking, and it’s clear that posts are still coming in with some frequency when the topic is hot enough.
By personal feed
Just for the heck of it, I also decided to track the posts per minute that appeared in my personal weibo feed. I only follow 71 people on weibo, so it’s a much slower trickle of posts, and I didn’t expect this data to be particularly useful, but I think it may have turned out more indicative of the reality than anything else I tracked. Here’s what I saw:
Click to view an interactive version of this chart.
Yup, I saw basically nothing. There’s really no pattern there at all. If I hadn’t been following the news and notified about it by Sina, I wouldn’t have even realized anything had changed.
Conclusions
I think that — nothing — is likely what most weibo users have experienced over the past few weeks. Certainly, they’re aware of the real name rules as some were forced to register, some were automatically registered and notified about it afterwards, and others had weirder things happen. But in terms of user experience, quantitatively, there doesn’t seem to have been a huge change. But of course, this is just a month or so of data from a very limited set of metrics. We’ll have to wait longer to see what the effects really are, especially if the anti-rumor campaign eventually leads to Sina being forced to implement a stricter (i.e. actually effective) real-name policy.
Qualitatively, it’s even more difficult to tell if things have changed. Has the quality of discourse altered? Are people being cowed into silence now that microblog providers theoretically have all their information on file? It’s hard to tell. Certainly the fact that rumors of a coup d’etat were being spread on weibo immediately following the real-name implementation seems to indicate the service’s political elements have not been scared into silence.
But it’s early yet, and everybody — Sina, the government, the users — are in this thing for the long haul. Let’s all keep watching.