Don't Try Too Hard to Please Twitter
January 22, 2015 9:22 AM Subscribe
The NYT Social Media team pulls the curtain back on how Twitter works for them with detailed examples of how changing text and descriptions and focus in their short messages resonated with readers, and which fell flat. Really interesting bit of transparency on their process, and results.
I had no idea they put so much thought into their tweets. Reading this, I was surprised at how they kept coming back to their philosophies on journalism or photo credits or other high level things that influenced small changes on their tweets.
posted by mathowie at 9:30 AM on January 22, 2015 [2 favorites]
posted by mathowie at 9:30 AM on January 22, 2015 [2 favorites]
OK, here's my question though. In the example of the security guard story, how many people didn't click through the second time because they'd already read the story? This doesn't seem to be good science.
posted by roomthreeseventeen at 9:48 AM on January 22, 2015 [6 favorites]
posted by roomthreeseventeen at 9:48 AM on January 22, 2015 [6 favorites]
A powerful entry from the Times’s Op-Doc series about a man going blind was illustrative of this problem and its solution. The first impulse in tweeting about the story was to highlight the series it was a part of, and the power of the audio diaries that were the basis of the video [they then got more clickthroughs when they stripped out the part about it being an audio diary]
I think they're really missing what happened here. It's not the "peacocking" that's a problem, it's the "it's an audio diary" that's the problem. I'll bet most of the extra people who clicked through the version that didn't own up to that still didn't listen to the audio diary. If you're browsing the web at lunchtime (or any other time) in your open-plan office or even if you're on your own and think "That looks interesting, I'll browse that to see if I want to read it in full" an "audio diary" is just a no-go area. I turn to the NYT because I want to read stuff; it's pretty rare that I'll ever click on one of their videos or audio links on purpose.
posted by yoink at 10:00 AM on January 22, 2015 [20 favorites]
I think they're really missing what happened here. It's not the "peacocking" that's a problem, it's the "it's an audio diary" that's the problem. I'll bet most of the extra people who clicked through the version that didn't own up to that still didn't listen to the audio diary. If you're browsing the web at lunchtime (or any other time) in your open-plan office or even if you're on your own and think "That looks interesting, I'll browse that to see if I want to read it in full" an "audio diary" is just a no-go area. I turn to the NYT because I want to read stuff; it's pretty rare that I'll ever click on one of their videos or audio links on purpose.
posted by yoink at 10:00 AM on January 22, 2015 [20 favorites]
Sara Critchfield, the editor of Upworthy, actually talks about the NYT in this video about optimizing the packaging of content for social media for clickthrough. I didn't rewatch the whole thing but it's either this video or another one on this page where she talks about how the Upworthy editorial staff generate 25 different headlines for every article and then select the best one. Say what you will about Upworthy but they've clearly put a ton of thought into this particular problem.
posted by skwt at 10:14 AM on January 22, 2015 [2 favorites]
posted by skwt at 10:14 AM on January 22, 2015 [2 favorites]
OK, here's my question though. In the example of the security guard story, how many people didn't click through the second time because they'd already read the story? This doesn't seem to be good science.
The NYTimes has ~14,900,000 followers on Twitter.
They look at their web analyitcs to see how many people clicked on the link and went to the story. It is probably suficiently below 15million to safely assume that they're exposing it to way more first-timers than already-clicked-ons.
posted by entropone at 10:16 AM on January 22, 2015
The NYTimes has ~14,900,000 followers on Twitter.
They look at their web analyitcs to see how many people clicked on the link and went to the story. It is probably suficiently below 15million to safely assume that they're exposing it to way more first-timers than already-clicked-ons.
posted by entropone at 10:16 AM on January 22, 2015
Avoiding the urge to “peacock” our work
I'd rephrase this as "No-one cares about your process, ever." The story is why people click through, not because it's part of a series.
This is a lesson I have to keep pounding into our own web folks. Content and matching what the visitors care about, first, last and always. Nobody gives a hoot about how we do things internally.
posted by bonehead at 10:22 AM on January 22, 2015 [13 favorites]
I'd rephrase this as "No-one cares about your process, ever." The story is why people click through, not because it's part of a series.
This is a lesson I have to keep pounding into our own web folks. Content and matching what the visitors care about, first, last and always. Nobody gives a hoot about how we do things internally.
posted by bonehead at 10:22 AM on January 22, 2015 [13 favorites]
This is an interesting FPP to come in to, thanks mathowie.
posted by infini at 10:34 AM on January 22, 2015
posted by infini at 10:34 AM on January 22, 2015
I'd rephrase this as "No-one cares about your process, ever." The story is why people click through, not because it's part of a series.
I care, a lot, about "process" to the extent of whether I'm clicking on a print story, a video story or an audio story. It really, really pisses me off to click on a headline and get delivered to a video--especially an autoplay video--without warning.
posted by yoink at 10:51 AM on January 22, 2015 [3 favorites]
I care, a lot, about "process" to the extent of whether I'm clicking on a print story, a video story or an audio story. It really, really pisses me off to click on a headline and get delivered to a video--especially an autoplay video--without warning.
posted by yoink at 10:51 AM on January 22, 2015 [3 favorites]
"Say what you will about Upworthy but they've clearly put a ton of thought into created this particular problem."
Click bait headlines are a race to stupidity. They reduce the amount of information I can get from a link summary, because everything becomes some variation on a very small set of manipulative emotional formulas. It's a poison.
posted by idiopath at 11:31 AM on January 22, 2015 [5 favorites]
Click bait headlines are a race to stupidity. They reduce the amount of information I can get from a link summary, because everything becomes some variation on a very small set of manipulative emotional formulas. It's a poison.
posted by idiopath at 11:31 AM on January 22, 2015 [5 favorites]
It isn't just @NYTimes; "ICYMI"-type tweets are common when linking to one's own content off-Twitter.
posted by mkb at 11:32 AM on January 22, 2015
posted by mkb at 11:32 AM on January 22, 2015
Click bait headlines are a race to stupidity.
@SavedYouAClick is actually quite good.
posted by colie at 12:28 PM on January 22, 2015 [1 favorite]
@SavedYouAClick is actually quite good.
posted by colie at 12:28 PM on January 22, 2015 [1 favorite]
ICYMI
"In case you missed it", for anyone wondering. (I had to go look it up.)
posted by a car full of lions at 12:39 PM on January 22, 2015 [3 favorites]
"In case you missed it", for anyone wondering. (I had to go look it up.)
posted by a car full of lions at 12:39 PM on January 22, 2015 [3 favorites]
There has never, ever, been a sausage I cared less about how it was made.
Click-thrus are like page views. The easiest things to measure matter the least.
posted by DigDoug at 12:58 PM on January 22, 2015 [2 favorites]
Click-thrus are like page views. The easiest things to measure matter the least.
posted by DigDoug at 12:58 PM on January 22, 2015 [2 favorites]
This makes me feel just a little bit better about journalism in the age of social media. Hoping my newsroom is paying attention.
posted by sixpack at 1:21 PM on January 22, 2015
posted by sixpack at 1:21 PM on January 22, 2015
I'd really like to see a similar discussion about the new "Watching" feature on the Times home page. It's almost like they're running an in-house Twitter competitor, a fast-updating aggregated feed of headlines, many of which aren't their own reporting. I've been finding it a pretty interesting idea and I'd like to hear from the people running it about their editorial perspective, and how they think the experiment is going.
posted by RogerB at 1:57 PM on January 22, 2015 [1 favorite]
posted by RogerB at 1:57 PM on January 22, 2015 [1 favorite]
I figured I got the same story tweeted to me day after day after day because they had a bunch of 19 year old interns who didn't give a shit running their Twitter.
No other reason ever occurred to me.
posted by sideshow at 2:03 PM on January 22, 2015 [1 favorite]
No other reason ever occurred to me.
posted by sideshow at 2:03 PM on January 22, 2015 [1 favorite]
I can't wait to see how The Times Is On It presents this.
posted by fedward at 2:25 PM on January 22, 2015
posted by fedward at 2:25 PM on January 22, 2015
I know that Twitter is pitching itself to brands as a real-time feed with however many metrics that back up how their target users look at the home page, but if you're the sort of person who actually follows their feed (you know, the sort of power user Twitter likes to pretend doesn't exist) then it becomes rapidly apparent that all the media brands are constantly A/B testing headlines and ICYMI-ing everything. After you click the third different headline for the same damn article you start to wonder if you should be following a media brand at all. I generally don't.
posted by fedward at 2:34 PM on January 22, 2015 [4 favorites]
posted by fedward at 2:34 PM on January 22, 2015 [4 favorites]
Click-thrus are like page views. The easiest things to measure matter the least.
You mean in that they are both the sole purpose and reward of participating online? The purpose of a website is to find an audience and be read, I'm not quite sure what utopian alternative you are proposing.
posted by kaspen at 2:45 PM on January 22, 2015
You mean in that they are both the sole purpose and reward of participating online? The purpose of a website is to find an audience and be read, I'm not quite sure what utopian alternative you are proposing.
posted by kaspen at 2:45 PM on January 22, 2015
They don't need a lot of followers, if their retweet reach is wide.
posted by mkb at 6:52 PM on January 22, 2015
posted by mkb at 6:52 PM on January 22, 2015
fedward: And it becomes impossible to follow some journalists on twitter because they're constantly doing the same thing (Ezra Klein and Matt Yglesias are god-awful about this. I hop on twitter decently frequently, and when you're posting five A/B test-y type tweets about the same article, and retweeting other people tweeting about it AND retweeting the Vox media twitter feed it's just annoying as hell).
Not everyone who is a journalist on twitter has this problem so it's sort of one more reason to be annoyed with those two, I guess.
posted by dismas at 7:08 PM on January 22, 2015 [2 favorites]
Not everyone who is a journalist on twitter has this problem so it's sort of one more reason to be annoyed with those two, I guess.
posted by dismas at 7:08 PM on January 22, 2015 [2 favorites]
For more transparency re: Twitter and other social media, here's NPR's Socialmediadesk on Tumblr. I've encountered no autoplay video nor audio on this tumblr for six months.
posted by Jesse the K at 7:51 AM on January 23, 2015 [1 favorite]
posted by Jesse the K at 7:51 AM on January 23, 2015 [1 favorite]
« Older An intersex perspective on gender critical ideas | Narcodrones Newer »
This thread has been archived and is closed to new comments
posted by ThePinkSuperhero at 9:26 AM on January 22, 2015 [1 favorite]