University of Missouri: tracking centerpiece performance
Background and process¶
Centerpieces get top billing on the homepage, are almost always shared on social media by noon each day, have at least one visual element and often have multimedia and supplementary links or sidebars, offering a richer experience for readers. Because we invest more editorial resources in these features, we wanted to get an idea of how this set of articles is performing against all other articles. Our hope is that they do much better.
In our content management system, Blox, we designate a centerpiece by giving them a “featured” flag. This metadata isn’t available in Parse.ly, so we had to also assign a tag “centerpiece” so that we could aggregate their statistics. Stories that were not tagged at the time of publication were tagged after the fact, and Parse.ly was able to pull those into our report retroactively. That said, it’s possible some centerpieces were not captured. A complete list of articles we tracked can be viewed here. Note that there are some duplicates, which we merged when we did our analysis.
As a baseline, we compared the set of centerpieces against all articles published during the same time period, roughly 1,000 total posts. We removed the Donald Trump piñata story from the calculation because it had such a dramatic effect on the averages of the overall post scores.
We looked at views, time on page, referrals, article length, multimedia and other factors to identify meaningful insights. To help with our analysis, we broke up the stories into four groups:
- Top 20 — Based on total views
- Bottom 20 — Based on total views
- Long reads — Ranked by time on page, focusing on those that had better-than-average time (1.4 minutes or higher).
- Social hits — Ranked by their social referral rate (% of visitors referred by social media).
Centerpieces outperformed overall content by nearly every metric on average, especially in engaged time. They had more visitors and views by double-digit percentages.
The biggest area where centerpieces lagged all other content was in search-engine referrals, receiving on average 46% fewer views from this source.
Centerpieces generated 5% more social referrals than the average article, a margin that has room for improvement.
There is no one type of story that dominated the top posts, but the most successful centerpieces tended to be timely and emotionally compelling.
Articles in the top 20 received 1,119 views or more. The bottom 20 received 254 or fewer.
Centerpieces in the top 20 received 32% of their visits from social. The bottom 20 averaged 13%.
Articles in the top 10 “long reads” had an average of 2.0 minutes of engaged time, well above the all-article average of 1.1 minutes.
The common denominator among long reads was that personality- or relationship-driven stories dominated this segment.
Articles that received more than 40% of their views from social also had almost double the total readership of the average centerpiece and almost triple the average article overall. Our CP’s should be able to perform really well on social, but not all of them do.
Takeaways and next steps¶
While centerpieces are outperforming the average on social, it’s uneven across all features. The outreach team should develop social sharing strategies behind each centerpiece prior to publication. We should focus our Facebook post writing for click-thrus and develop better second- and third-day sharing techniques, as many of these articles have long shelf lives. In the spring, outreach should experiment with audience targeting tags (a new Facebook feature) to improve reach.
Search-engine traffic is extremely lagging for centerpieces, so a renewed focus on SEO-friendly headlines could make a significant difference in traffic to these stories.
Consider compiling CP’s into a collection or series of collections based on themes that can help readers discover content that might appeal to them. This collection could be a hub page on the website or a featured block that stays on the homepage. Most of our CP’s have a long shelf life, making it easier for them to be discovered by a variety of readers can build readership of an article over time.
The ability to track analytics performance based on tags is a key feature of Parse.ly that could be further explored. For example, we could tag specific beats or topic areas that we want to track and better understand. A story that has a video or a special embedded feature could be tagged, so we can understand how those elements affect engagement.
2018 traffic sources by content categories and topics
An analysis of referrer traffic data reveals benchmarks for the categories and topics that resonate with 2018 audiences and where they discover stories to read.
Ahead of the 2018 midterms, a look back at the 2016 presidential candidate media coverage
Turning to data to understand the full story of how the media covered the 2016 elections and how readers responded.