Tom Webster, writing and speaking

Grabbing Headlines and Survey Reporting

Added on by Tom Webster.

The trend of awful Twitter stat reporting continues. This time, it's AdWeek telling us that "19% of U.S. Internet Users Tweet" and that "the army of Twitterers is growing quickly." Makes for a great headline, but again, not borne out by the actual survey. What the Pew survey asks, specifically, is whether or not Americans update their status on services like Twitter. The question wording, in fact, is "Twitter or another service to share updates about themselves or to see updates about others." One thing that is clearly going on here is the fact that "Tweet" is becoming like "Kleenex" or "Band-Aid"--fast becoming the generic term for short status updates, even on competing services.

I continue to despair at the state of journalism when it comes to data like this. I am already seeing this crop up on sites as "19% of Americans [not even Internet users] Tweet!" A pocket calculator would tell you that 19% of Americans would be about 60 million people. There is plenty of clickstream data out there to make the point that this is a patently ridiculous estimate of actual Twitter users. Yet I will see 19% reported--and retweeted--all week long. But maybe it's time for me to stop beating up on journalists so much, and look at the actual survey in question.

The last time I quibbled with the reporting of this data, I pointed out that most people actually update their status with Facebook, not Twitter, and showed the traffic as a gross comparison. The gap between Facebook and Twitter continues to widen, if anything, making the wording of this particular question curious, indeed. While I will continue to highlight the improper reporting of stats like this, maybe it's time to challenge the wording of this question itself. Asking Internet users if they use Twitter "or another service" to update their status is roughly akin to asking them if they consume "Sprite or another soft drink" and reporting the results as "Sprite Drinkers." I realize that this would blow up the trending on this particular question, but if Pew themselves are going to continue to call this report "Twitter and Status Updating," it would be helpful for the rest of us if they would break the question out into "Twitter status updating" and status updates on other services.

The resultant data would be less sensational, perhaps, but far more useful. It is inevitable, as status updating apps proliferate, that the role of Twitter in status updating will continue to shrink, not grow, while the Pew number goes up with each report. As I write this, I still have no idea how many unique Americans are actually using Twitter, and that in itself would make for a sensationalist headline, no?