ellipsis flag icon-blogicon-check icon-comments icon-email icon-error icon-facebook icon-follow-comment icon-googleicon-hamburger icon-imedia-blog icon-imediaicon-instagramicon-left-arrow icon-linked-in icon-linked icon-linkedin icon-multi-page-view icon-person icon-print icon-right-arrow icon-save icon-searchicon-share-arrow icon-single-page-view icon-tag icon-twitter icon-unfollow icon-upload icon-valid icon-video-play icon-views icon-website icon-youtubelogo-imedia-white logo-imedia logo-mediaWhite review-star thumbs_down thumbs_up

How time of day affects content performance

How time of day affects content performance Ky Harlin

Take a look at these three stories:

What you may notice that in both subject and spirit, each article is distinct from the others -- a combination of the funny, shocking, informative, and awe-inspiring. But what do they have in common?

To start, they represent the top BuzzFeed articles in terms of the amount of views received in a single hour over the past year. In other words, these three articles saw the biggest spikes in traffic of anything we've published recently. But the similarities don't end there -- these spikes all occurred within the same two-hour period of the day!

There's always been the sense among publishers and marketers that content is key, and I don't disagree with that. Each of the examples above owes the majority of its success to the fact that people found the content interesting and sharable. But the fact that each example took off at the same time of day also illustrates a pivotal role that time of day plays.


On the surface, it seems obvious -- it's well known that web traffic fluctuates in correspondence with time of day. People browse the web more frequently during certain hours of the day. In a general sense, this is easy to understand. For example, people are sleeping at night, so they tend not to browse the web at night. But for content creators, understanding the intricacies of these fluctuations is essential. Recent BuzzFeed data reveals a highly predictable daily pattern of web traffic where internet users reach their peak at in the early evening. (Note: My analysis looked at hourly web traffic to posts on BuzzFeed and across our partner network, a total of over 300 million monthly visitors. The data goes back to April 2011. To account for time zones, I looked at U.S. traffic only and to account for seasonal or daily fluctuations, I cross-validated my observations over various date ranges.) What follows is an explanation of this pattern and information on what publishers and marketers should be doing to account for it.

The pattern

The plot below demonstrates the total number of page views during each hour of the day.

Before we begin analyzing this line, let me note: regardless of the time range you choose or the day you're looking at, its shape remains remarkably similar. The peaks and troughs may shift slightly depending on other qualities, but the overall shape remains the same.

Now, let's begin examining this patterns of how page views fluctuate throughout the day. Here are the basics:

  •  A daily low point occurs in the early morning

  • This low point is followed by a steady, almost linear period of growth until lunch time

  • This growth slows during lunchtime where we begin to see a plateau

  • Following lunch hours, another hour or two of modest growth occurs

  • During the mid-to-late evening, traffic drops slowly but steadily

  • As we approach the end of the day and the beginning of the next, traffic plateaus

  • Starting around the beginning of the next day and continuing until the early morning, there is a steady, almost linear period of decline until we reach the morning low point

This pattern probably won't come as a shock to many of you. It's easily explainable given the normal trajectory of one's day. What might be more surprising is the high predictability of the pattern -- over the 1,700 days I looked at, 86 percent follow the above pattern. Furthermore, this is independent of extraneous factors, such as the number of articles published on that day.


Despite this highly predictable pattern, publishers don't seem to be accounting for it as much as they should. Let's look at this same graph, but add a green line showing the number of articles published during each hour of the day.

The time of day that articles are published shows little correlation with the hours of highest traffic.  Essentially, articles are published at one constant rate throughout the working day and another constant rate from the early evening to early morning, with the former rate about 10 times higher than the latter.  Given that traffic is not consistently high during the hours of high production, a lot of content isn't given the proper chance to "seed." In other words, lots of content isn't shown to a sufficient number of visitors to be reach the full potential to spread organically. Thus, it's likely that a lot of content that's published at, for example, 9 a.m., doesn't receive much traffic in the first hour and is hence removed by an editor, would have taken off if published at 3 p.m.

Publishers are marketers and need to realize that content performance doesn't scale linearly -- judging performance by the reaction of a small subset of your visitors is often flawed.  Adjustments should be made based on the times of day that people most frequently visit the sites of interest. If there's not a lot of people on your site at a certain hour, don't publish as much -- let the stories you have stick around a little longer and see if they take off.  And if there's a lot of people on the site during another hour, publish often -- the stories will be seen by large numbers of visitors which gives you the chance to try many different pieces of content to see what sticks.

Ky Harlin is a data scientist at BuzzFeed, Inc.

On Twitter? Follow Ky at @KyHarlin

Follow iMedia Connection at @iMediaTweet.

Ky Harlin is a data scientist with over six years of experience in data-driven research, mining, and software development.  He is currently the data scientist at venture-funded New York start-up BuzzFeed, where he studies viral content under...

View full biography


to leave comments.

Commenter: Andy Cheng

2012, April 24

I think this article covers the general public web usage well. On another hand, if your content is business oriented, like us, the traffic is super high in the morning and gets lower toward later part of the day.

Commenter: Nick Stamoulis

2012, April 24

I'm actually surprised at how high engagement is so late in the day. I would have expected it to drop a lot faster, especially around dinner time. I guess businesses need to start scheduling content to go out after hours!