Max, great analysis. I was curious if you researched whether or not
BuzzFeed is using the tactic of multiple titles for a single
(partially generated) article?
The multi-title tactic was mentioned in the "The King of Clickbait" article and discussion:
Fascinating topic, but nothing surprising about any of the findings. (Does that qualify this post as clickbait?:))
An extremely interesting question IMO would be what is the longetivity of these tactics? Are there phrases that worked extremely well for a period of time, and then stopped being effective? In other words, how long before people get tired of clickbaits?
> An extremely interesting question IMO would be what is the longetivity of these tactics? Are there phrases that worked extremely well for a period of time, and then stopped being effective? In other words, how long before people get tired of clickbaits?
I mentioned at the end at the disenchantment against linkbait. Linkbait works due to basic human psychology, similar to ancient psychological marketing tricks like odd-even pricing and price anchoring.
Unless basic human psychology changes, linkbait will still work. There's still a negative brand perception for BuzzFeed though; it wouldn't surprise me if they attempt a rebrand at some point if they want to be a serious publication.
> In other words, how long before people get tired of clickbaits?
Somewhat facetiously, "is clickbait more like banner ads, or more like porn?" Is it useless, or is it something people actually want sometimes? Once people realized that flashing "punch-the-monkey" gifs never led anywhere useful, they quickly stopped clicking. However, porn is something people actually want, and it just needs to be "good enough," not "the very best porn on the internet."
While clickbait is low-quality content, it's still useful to someone looking for a 10-minute distraction while on the bus or at lunch. Similarly, Demand Media's content farms are extremely low quality, but might be good enough if I'm just searching for how to brine a turkey or something. Like soap operas and 24-hour "news," I think clickbait serves a real need, and isn't going anywhere.
The fallacy of the clickbait charge against Buzzfeed is that they don't use clicks as a metric, they focus on shares. Buzzfeed learned from the once hot Upworthy (who originated the "You Won't Believe What Happened Next" style headline) that if a visitor feels like they were tricked into clicking on a link to your site then they aren't going to share it. You get a lot more traffic by delivering on the promises made by your headlines.
I think it's kindov subjective. A clickbait headline may "trick" a user into clicking on a link but it may also deliver on the expected content.
Is there a blog post dissecting above hypothesis though. I'd be interested in reading it.
I'll tell you guys a funny story. So, the Onion has this site: http://clickhole.com Check it out, it's great parody. I actually go there.
So, the last time someone mentioned BuzzFeed - which I don't really go on, I don't recognize it or anything - I ended up opening it and opening Clickhole.com in adjacent tabs.
Now, since I don't read buzzfeed, I was then completely convinced it was clickhole. It had like articles like this (from the current front page, not what I was reading but they'll do):
-23 Ways You Know Your Obsession With Your Cat Is Getting Out Of Control
- The Definitive Ranking Of The Most Cringingly Great Dialogue In “Titanic”
and so forth. (The above are verbatim from the site right now.) But then I got to one, and I couldn't figure it out. (Just looking at the title.) I've actually found it just now, based on remembering a keyword. It was:
- "Muslim-Owned Shops In Birmingham Were Attacked With Guns And Hammers"
I was looking at it and looking at it, trying to figure out the "joke" (why guns and hammers.) It was like a pun I couldn't get. Like the Onion was too clever for me. After about a minute, I gave up. And I saw on the tabs that I was on an actual site, not clickhole. I had to just laugh and laugh. The reason I couldn't figure out the joke was that I was on buzzfeed, which isn't satire! I just couldn't stop laughing.
I'm teaching a Unix/data course this quarter, and inspired by the OP (who I believe did an analysis/visualization prior to this) and other BuzzFeed-title analysts, I put together an exercise in doing a frequency distribution of BuzzFeed titles (really, an exercise in basic HTML parsing and regexes) http://www.compciv.org/homework/assignments/buzzfeed-listicl...
I gathered the links through their archive pages, which go back to 2006...I didn't run the exercise yet myself, but at one point did a quick count of total titles since 2006 and counted more than 200,000. That didn't seem right (and maybe my selector was off)...but the other day, I tried it with just the 2014 year and came up with more than 60,000 articles.
...that sounded clearly off (the OP said he scraped only ~60,000 distinct articles)...but when I visited one of the archive pages to eyeball it...yeah, 60,000 articles in a single year looks about right: http://www.buzzfeed.com/archive/2014/11/18
It's even more astonishing when you go back to the 2006 pages and see that there were days/regular weekends in which BuzzFeed published no content at all.
Huh, I was not aware of the archive at all. However, looking at the archive, it appears that there are a lot of community, user-added posts so I'm unsure of the reliability of those posts.
Yeah it took me a bit of time to find the archive. Unlike you, I have a little more faith in BuzzFeed's SEO expertise and figured they would have to have some kind of sitemap ;)
Don't know about the community posts or how they're structured with the main content...it might be possible on a second pass to filter out non-Buzzfeed-staff posts based on the URL (which I assume contains the username of the author).
(Also, I don't know if they changed this since you attempted your last scrape via pagination...but now if you paginate too far, you will be rewarded with a Boyz II Men video)
The analysis is right on. I once experimented with creating a single page website, literally a single page, by posting a listicle with photos. The webpage got over a million shares on Facebook.
> Additionally, BuzzFeed was one of the first news sources to use non-neutral headlines that deliberately invoke a reaction in the reader which then subsequently tempts them to click on the article in an attempt to promote virality.
I don't think this is a new phenomenon. Yellow newspapers like The Daily Mail have existed since shortly after the invention of the printing press.
What I would love is a fully-integrated browser extension that detects/fades/hides clickbaity headlines/links/articles, using something like Bayesian spam filtering. I know there's already a node.js package that worked with this concept (https://github.com/TJkrusinski/clickbait), but it would be great to have something that works like Adblocker on the various CSS classes where headlines usually lurk, something that would check all headlines instead of just some sort of blacklist for a few certain sites. Maybe it would include a crowdsourcing aspect, where users would, given two headlines, choose which one is more clickbaity, to train the system.
23 comments
[ 3.0 ms ] story [ 37.5 ms ] threadThe multi-title tactic was mentioned in the "The King of Clickbait" article and discussion:
https://news.ycombinator.com/item?id=8810670
http://www.newyorker.com/magazine/2015/01/05/virologist
The dupe articles had the same exact title, however.
An extremely interesting question IMO would be what is the longetivity of these tactics? Are there phrases that worked extremely well for a period of time, and then stopped being effective? In other words, how long before people get tired of clickbaits?
I mentioned at the end at the disenchantment against linkbait. Linkbait works due to basic human psychology, similar to ancient psychological marketing tricks like odd-even pricing and price anchoring.
Unless basic human psychology changes, linkbait will still work. There's still a negative brand perception for BuzzFeed though; it wouldn't surprise me if they attempt a rebrand at some point if they want to be a serious publication.
Somewhat facetiously, "is clickbait more like banner ads, or more like porn?" Is it useless, or is it something people actually want sometimes? Once people realized that flashing "punch-the-monkey" gifs never led anywhere useful, they quickly stopped clicking. However, porn is something people actually want, and it just needs to be "good enough," not "the very best porn on the internet."
While clickbait is low-quality content, it's still useful to someone looking for a 10-minute distraction while on the bus or at lunch. Similarly, Demand Media's content farms are extremely low quality, but might be good enough if I'm just searching for how to brine a turkey or something. Like soap operas and 24-hour "news," I think clickbait serves a real need, and isn't going anywhere.
Pah. There are numerous interesting findings in there.
1. Odd numbers outnumber even numbers, except for "10"
2. The optimal number of items in a listicle is 33-45.
3. "There’s a strong focus on nostalgia, with toys, childhood, and 80s. "
4. "Gay" is surprisingly common keyword
5. "Which" and "Character" get some of the highest FB shares. (possibly from "Which character are you" quizzes)
> longetivity of these tactics
The first bar graph shows Listicles gaining ground on normal articles until late 2013, then stabilizing over the last 15 months.
So, the last time someone mentioned BuzzFeed - which I don't really go on, I don't recognize it or anything - I ended up opening it and opening Clickhole.com in adjacent tabs.
Now, since I don't read buzzfeed, I was then completely convinced it was clickhole. It had like articles like this (from the current front page, not what I was reading but they'll do):
-23 Ways You Know Your Obsession With Your Cat Is Getting Out Of Control
- The Definitive Ranking Of The Most Cringingly Great Dialogue In “Titanic”
and so forth. (The above are verbatim from the site right now.) But then I got to one, and I couldn't figure it out. (Just looking at the title.) I've actually found it just now, based on remembering a keyword. It was:
- "Muslim-Owned Shops In Birmingham Were Attacked With Guns And Hammers"
I was looking at it and looking at it, trying to figure out the "joke" (why guns and hammers.) It was like a pun I couldn't get. Like the Onion was too clever for me. After about a minute, I gave up. And I saw on the tabs that I was on an actual site, not clickhole. I had to just laugh and laugh. The reason I couldn't figure out the joke was that I was on buzzfeed, which isn't satire! I just couldn't stop laughing.
I gathered the links through their archive pages, which go back to 2006...I didn't run the exercise yet myself, but at one point did a quick count of total titles since 2006 and counted more than 200,000. That didn't seem right (and maybe my selector was off)...but the other day, I tried it with just the 2014 year and came up with more than 60,000 articles.
...that sounded clearly off (the OP said he scraped only ~60,000 distinct articles)...but when I visited one of the archive pages to eyeball it...yeah, 60,000 articles in a single year looks about right: http://www.buzzfeed.com/archive/2014/11/18
It's even more astonishing when you go back to the 2006 pages and see that there were days/regular weekends in which BuzzFeed published no content at all.
Don't know about the community posts or how they're structured with the main content...it might be possible on a second pass to filter out non-Buzzfeed-staff posts based on the URL (which I assume contains the username of the author).
(Also, I don't know if they changed this since you attempted your last scrape via pagination...but now if you paginate too far, you will be rewarded with a Boyz II Men video)
https://chrome.google.com/webstore/detail/baitblock/nnonoddk...
https://chrome.google.com/webstore/detail/gawkblocker/fbdakf...
Now now, let's not go overboard. BuzzFeed is a new tabloid, they didn't invent tabloids.
I don't think this is a new phenomenon. Yellow newspapers like The Daily Mail have existed since shortly after the invention of the printing press.
Unfortunately, it's sad to think that this is the state of journalism. <sigh>
I built a search engine that ranks on social shares and the social comments.
It's great for avoiding click bait in results.
http://engu.me