Yahoo! Labs Study Has Surprising Findings

A heatmap of The New York Times' most popular headlines' buzzwords. The bigger the word, the more stories the word is included in while the redness of the word indicates the number of likes associated with the word.

A heatmap of The New York Times' most popular headlines' buzzwords. The bigger the word, the more stories the word is included in while the redness of the word indicates the number of likes associated with the word.

A study from Yahoo! Labs has been published which uses Facebook likes to better understand the dynamics behind online news media. Of the many findings reported in the study, the most surprising ones included:

  • News does not have to be updated frequently to be popular.
  • Op-Ed isn’t dead.
  • Flies and online articles have similar life spans.

Publishing Frequency

Many editors are adamant about the frequency of publishing articles and feel that this is vital to maintaining a reputation of being a reliable news source. The study, however, does not necessarily reflect this sentiment. Three out of the the five most “liked” news sources only update articles once a day (these being NPR, The Guardian, and Yahoo! News with rankings of 3, 4, and 7, respectively).

I hesitate, however, to passively comply with this finding; perhaps there is more that could be speculated about it. While the data shows that infrequently-updated sites are doing well with Facebook likes, it is also important to think about the people liking these pages who actually influence the statistics. While growing in popularity with adults, Facebook is still used by younger people, who are not traditionally as invested in reading the news as older people are. Thus, it is important to emphasize that most people liking articles are not looking to read pages amongst pages of writing, but perhaps liking something because it caught their eye instead. There may not be a great personal investment in the newspaper itself which drives a person to come back to a news site; instead, what may be driving these likes for sites such as NPR may be its share-ability factor: while the Times has the largest audience to amass likes from, NPR may have more interesting headlines which readers may want their friends to know about on Facebook.

Importance of Opinions

Out of a list of 40 top-liked articles, only four were about politics. Even more surprisingly, only three were from the opposite spectrum: celebrities. The most common top story pieces were actually opinion articles, with the top article being “Why Chinese Mothers are Superior” in The Wall Street Journal.

This is a finding which especially surprised me considering the talk our class had with Paul Bass of The New Haven Independent a few weeks ago who said, “Op-Ed is dead.” Nor is he the only one to believe this either, as more and more op-ed columnists are fearing the loss of their jobs. Rather than mix two groups together, a more succinct way of describing this phenomenon is by separating online and print opinions into different realms. In these terms, we can better understand this supposed paradox: print media on a whole is on the decline, but it’s obvious that online blogging and citizen journalism are only two sources from which opinionated pieces can stem.

Life of an Online Article

Like that of a fly, the life of an article seems to last only 24 hours. After that 24-hour time period, the number of Facebook likes an article receives plummets. On average, over 80% of Facebook likes happen in the first day despite the search-archive feature that nearly every news site provides.

Why is this surprising? While I understand that a good amount of publicity is bound to happen, I did not think that the likes would plummet at such a drastic rate after just a day. The power of information cascade seems as though it would disprove this theory: if 100 people like the story today, then those 100 people will have shown it to 100 different networks of people who will then show it to an even larger set of people. It seems contrary to this that the publicity wouldn’t slowly decrease over time.


While I have my doubts as to the overall concreteness of the study, I do think this analysis of Facebook likes is crucial to understanding how exactly online news travels through social media networks. When retrieving data from a very specific source, one cannot merely use this type of study as standalone proof of online news distribution patterns, but it is certainly not a study that should be ignored, either.

This entry was posted in National News. Bookmark the permalink.

3 Responses to Yahoo! Labs Study Has Surprising Findings

  1. Colby Brown says:

    Great post, Raquel. This shows that the criticism I posted of the New YOrk Times is indeed an Internet wide-phenomenon. To me, the fact that the popularity of Internet content depends so much on timeliness seems outdated, arbitrary, and broken.

    It amazes me that the technology for delivering online content has advanced so greatly, and yet the way it is sorted, organized, and prioritized remains so unsophisticated. To prevent the death of important news-stories, I think that content providers should create a distinction between top news and latest news – making efforts to convey the importance of both.

    The architecture of Facebook and the patterns of “Facebook likes” are a microcosm of this problem. As time goes on, the news you are likely to see is limited to ‘the content that your friends have read most recently, among a pool of recently posted articles’.

  2. Shunori says:

    This is a really interesting study – thanks Raquel! I agree with Dilan that the structure of Facebook often defines the lifetime of an article (especially in this situation, where the lifetime is being measured more by “likes”). I would also argue, though, that the 24/7 news cycle that the internet has exposed us to has meant that there is a constant flow of information. Stories become old so quickly (unless they are particularly important/catchy, in which case I’m sure they live on) – we are constantly playing catch up. I would think that that contributes to the short lifetime of articles.
    Also, to the point that you raised about publishing frequency and the number of “likes” NPR, the Guardian, and Yahoo! News are getting – I wonder if that has anything to do with their paywall systems (compared to their competitors)?

  3. Devin says:

    While this study may reflect some insights into online news distribution, I think it also says a lot about the way Facebook works. If someone were to share a news story on Facebook, they’ll post it and that post will appear on their friends’ mini feeds for maybe the next two or three hours. After that time, newer news has overwhelmed the “old” post and the only time a friend will see that post is if they were included specifically in the original post or if they go to the original person’s Facebook profile page. Therefore, after 24 hours, old posts are seriously old news and are buried. The conclusions concerning the time-frame for sharing news from this study may say a little about the actual spread of the news, but it definitely says a lot about how quickly Facebook requires its users to keep up. The news stories themselves may hold value after 24 online, but the “likes” on Facebook drop off due to Facebook’s architecture as opposed to the value people place on the news stories.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>