We wanted to know: Does posting more increase your overall traffic (page views) or is there a point at which there is simply too much content and the reads per post starts to drop off?
Here are the data behind that somewhat simplified graph above. We looked at publishers in our system who have the Repost button on their content pages, which allows us to track both how much content they post and how many views each post gets in its first week. Then we scaled the data so that the median post equaled 1000 views (this allowed us to compare sites with wildly different traffic levels). Once we had that data, we plotted the number of posts in a week vs. overall traffic those posts generated in their first week. While major external events (elections, bombings, etc.) generated outliers, the overall pattern was remarkably clear.
Sites that post more than 5 articles a day tend to have significantly more traffic than those that post fewer than 5 articles a day. While it is not possible to tell if this is cause or effect from the data we have, we can say that among sites of any size during weeks when those sites posted more content, the traffic scaled with the number of articles.
Here is the result for original (native) content:
Here is the result for articles reposted from other sites:
The trend lines are 2nd order polynomial regressions. For both original and reposted content, there is a strong correlation between the frequency of posting and overall site traffic. Put simply: more content = more traffic.
Overall the median reposted article gets 86% of the traffic that the median original article gets. In other words, the source of the content plays a minor role in how well it does. Editorial voice, the choice of what to publish, is more important than authorship.
- We tracked page views not unique viewers. We don’t have the data to reliably predict how an individual reader will respond to increased frequency. Anecdotally, we know that when people return to a site that hasn’t been updated, they tend not to come back as often (or at all); however, we don’t know how that behavior changes once the update frequency increases beyond five posts per day.
- We only tracked content that was available for syndication or syndicated via Repost, so content syndicated by sites from other sources (eg AP, Reuters, etc.) is not included.
- We used data from the first week after the posting of an article because 50% of the traffic happens within the first 23 hours after posting and search-related long tail traffic is pretty much random and doesn’t correlate with post frequency.