Michigan State University

Michigan State University’s The State News Plans Section Content Based on Data

The State News, at Michigan State University, has been using Parse.ly since 2015 to figure out what types of content drives readers to its site, and why. See here and here for past analysis.

Fall 2016: Project Overview
Jake Allen

Editor in Chief, The State News

Editors at The State News analyzed the data obtained from articles published between November 30, 2015 and November 30, 2016 to determine which received the most engagement by section (City, Campus, Sports, and Features).

Each section was divided into specific categories of coverage, including breaking news, event coverage, and in-depth reporting. Understanding the traffic patterns to each section (and category of coverage) provided editors with valuable insight into the types of content to pursue in the Spring 2017 semester.

Results

Sports was the most popular section at The State News during the project period — receiving the most views and visitors of any other section. However, stories posted within ten minutes of a game ending never appeared among the top 50 posts in the Sports section. So, editors have decided to focus on more popular categories of Sports posts, including sports features and interesting athlete profiles in Spring 2017.

Breaking news articles were the top-performing category in the Campus section of The State News, so the editorial team will double down on producing this type of content. According to Jake Allen, The State News Editor-in-Chief: “If a newsroom employee sees breaking news anywhere in the MSU or East Lansing community, whether you are a reporter, photographer, editor, or designer, you need to inform one of the editors and help coordinate efforts to get the story covered.”

In the City section, event coverage stories made up the majority of the top posts for the semester; however, the editorial team realized it could do a better job covering breaking news in this section. As a result, the newsroom has vowed to pay more attention to local papers and build better relationships with the East Lansing Police Department and other emergency responders in Spring 2017.

Editors noticed a drop in traffic in the Features section of The State News in Fall 2016, so they decided to focus on writing personality profiles (which have consistently performed well) next semester. They plan to avoid event coverage, which was not as successful in terms of number of visitors.

Show less

The University of Virginia

Creating a Digital Publishing Strategy: A Video by The University of Virginia's Pioneer

The Pioneer created a short video about their approach to analytics and using Parse.ly to hone in on a publishing strategy that worked for them.

Creating a Digital Publishing Strategy: A Video by The University of Virginia's Pioneer

The Pioneer Staff

 
Show less

The University of Virginia

Analyzing with Reports: Using the Parse.ly Reports Feature to Create Content Strategy

The Cavalier Daily delves into the Parse.ly dashboard, discovers what engages their readers, and uses that knowledge to plan for the future of the publication.

Analyzing with Reports: Using the Parse.ly Reports Feature to Create Content Strategy
Dani Bernstein

Editor-In-Chief, The Cavalier Daily

After testing out Parse.ly’s analytics in the fall, this spring we used Parse.ly’s report-generating function to compare our analytics data over time and find the best ways to improve our page views and readers’ on-page duration. The reports function — which creates easy-to-read, visually appealing reports in seconds, based on whatever customized factors we want — has quickly become essential to our long-term operations.

Last semester, our concern was a drop in online readership — something we still monitor. My predecessor noted how we had editors track which modes of presentation were most effective, what news merited more coverage, and where our referrals were coming from via Parse.ly beginning in September to combat this issue. This semester, however, with our enhanced knowledge of analytics from the fall, we took our use of Parse.ly a step further, focusing mostly on readers’ duration on the page and whether they moved between articles after being referred to our site.

What we discovered was that duration on the page was heavily correlated with the presence of multimedia — including graphics, videos, photos, timelines, charts, you name it — because our readers ultimately want a visual product. The same characteristics that drew us to Parse.ly — its simple, aesthetically appealing, and informative platform — were keeping our readers on some of our articles, while a lack of visual appeal deterred them from staying on others. We doubled our efforts to incorporate multimedia into our work, expanding our multimedia teams, thereby making room for more students to join our newspaper in the process.

Moving forward, as we plan a website redesign for our paper, we hope to make our site as visually appealing as we’ve now learned it needs to be. We would not have understood that need — what our readers are most drawn to — were it not for Parse.ly’s analytics. In order for us to provide our peers with the best news and in the most accessible way, we need to offer that news in the best possible format. We learned that from using Parse.ly, a now-essential tool for our newsroom.

Download the PDF Report

Show less

Texas Christian University

Lessons Learned: TCU’s JOUR 30204: Reporting Class’s Observations and Findings

Students in TCU's JOUR 30204: Reporting are chief contributors to the TCU 360 site. As part of their coursework, they were asked to analyze their individual work using Parse.ly. For most, it was their first ever exposure to an analytics tool.

TCU 360 Parse.ly Report
Bob Schieffer

TCU used their Parse.ly access to gain insights into engagement duration, social media amplification, mobile views, popular content, and more! Their project was completed as a slideshow, which you may download to read their full findings.

tcu_pdf
Show less

Michigan State University

Bonkers and Clunkers: Identifying Both High and Low Performing Posts

Michigan State University is using Parse.ly to better better understand its audience and tailor posts to days and times when they have the best chance for engagement.

State News Newsroom Uses Parse.ly
Robert Hendricks

State News Editorial Adviser

Using Parse.ly this past semester has allowed our newsroom to more closely monitor what our readers want and more importantly, how they want it. Because we were able to adapt to what our readers like — such as shorter articles, more visuals, and more easily readable news snippets — our page views were on average higher than they had been last year.

Parse.ly allows us to better understand our audience and tailor our posts to days and times when they have the best chance for engagement.

One of the more useful features of Parse.ly was the ability to run custom reports. Each week we not only looked at what was most popular, but also looked at what posts were the least popular. During our fall training, Sarah Frank, a former State News student journalist and current executive producer at NowThis News, talked to us about how they rate their stories into two categories - Bonkers (posts that do well) and Clunkers (posts that do poorly). She said it was important to identify the least popular posts to identify why they did poorly and how to improve in the future.

We started this same type of report at the State News and the newsroom really responded to it.

In this particular case 5 of the 10 Clunkers were published on a Friday. Which is typically not a high traffic day for the State News. But also the type of stories were not particularly the most engaging either for those five Friday stories.

Parse.ly allows us to better understand our audience and tailor our posts to days and times when they have the best chance for engagement.

Overall, Parse.ly helped us do what many other news organizations are constantly trying to do — figure out what our readers really want, and giving them more of that, while trimming the fat on working to produce stories that might not be as popular. I know too that giving the students an opportunity to use a professional level product will help them as they are looking for jobs after graduation.

Show less

Michigan State University

Stating the Facts from Michigan State’s Spring Semester

See how Michigan State University used the metrics from Parse.ly to inform their news coverage in a downloadable resource.

Download the State News PDF

University of Virginia

How The Cavalier Daily Made the Most of Parse.ly Analytics

Concerned about a drop in online readership, The Cavalier Daily, University of Virginia’s independent student newspaper, is now using Parse.ly to actively engage with its site analytics — helping editors to make more informed decisions based on data.

How The Cavalier Daily made the most of Parse.ly analytics
Julia Horowitz

Editor-in-Chief, The Cavalier Daily

My generation consumes digital media at astronomically high rates. It often comes as a surprise, then, when I tell people that one year ago The Cavalier Daily — the independent student newspaper at the University of Virginia — had only a rudimentary understanding of web analytics. Producing more than 100 pieces of online content per week, we had a wealth of information at our fingertips. But we were doing little to take full advantage of it.

When we returned to school this fall, however, readership began to drop — at times by 20 percent. We quickly realized our online product was suffering because we weren’t using the right tools to analyze our audience. Top editors had access to data on Google Analytics, but struggled to find time between classes to synthesize the information into something useful. Our home page remained static throughout the day, and no one was critically examining what types of content resonated most with digital audiences. We claimed to be a digital-first newsroom, but weren’t living that reality.

The platform is both easy to use and highly informative — a necessity if busy college news editors are expected to make full use of the tool.

To combat this problem, we started using Parse.ly in September. We gave all 26 literary editors access to the platform, working to cultivate a culture of digital thinking across the newsroom. At the end of each week, we asked that editors use the reports feature to evaluate popular content, from there noting which modes of presentation were particularly effective and which subjects merited further coverage. Meanwhile, our social media team began to closely watch referral traffic, looking for the peak times to promote content and which social media sites reached the majority of our readers.

The platform also helped us engage with our content on a more granular level. The Cavalier Daily home page became a more fluid space as we observed which stories the community cared about most throughout the day. A story that was gaining particular traction via Facebook could be pushed out a second time that evening. When we had a slow day, we could finally say why. We even used Parse.ly to help generate new content, identifying and aggregating the most popular stories of the term for our “Semester in Review” issue in December.

Our newsroom’s Parse.ly launch marked the first time many of our editors actively engaged with analytics and used the information to make decisions. They’ve reported back that the platform is both easy to use and highly informative — a necessity if busy college news editors are expected to make full use of the tool.

With Parse.ly’s help, we’ve begun to foster the digital-first culture essential for running a successful newsgathering operation in 2016. Not only has readership stabilized; it is now climbing.

This much is clear: to continue to provide our university community with information that is new, relevant and insightful — as is our mission — we need to be where our audience is. And for us, Parse.ly has been part of the answer.

Show less

Claremont McKenna College

Understanding Their Audience: The Golden Antlers’ first Foray into Data Analytics with Parse.ly

Claremont McKenna’s student newspaper, The Golden Antlers, is using Parse.ly to provide insights and analysis into optimization patterns — including which day or time is best to release an article, what topics resonate most with their audience, and more.

The Golden Antlers and Parse.ly
Jackson Smith

Managing Editor, The Golden Antlers

The Golden Antlers is a campus satire publication for Claremont McKenna College, focusing on exposing the contradictions and hypocrisy inherent in modern college culture in a humorous manner.

Parse.ly gave our publication access to a set of deep, rich analytics that helped gauge reaction to our articles, inspire our authors to write more effectively, and shape the future of our organization.

Parse.ly became the compass for The Golden Antlers, providing key insights to help guide us out of the dark in analytics.

From the outset, Parse.ly blew Google Analytics out of the water. On the very first screen our team sees when opening up Parse.ly — ­­the “Overview” tab — we are able to, at a glance, gather an enormous amount of useful information. Without clicking anywhere, we are able to see whether or not the current day is performing at, above, or below usual (a process that used to take a series of manual calculations) and w​hich ​posts are receiving the most attention. Immediately after releasing a new article, we are able to see whether or not it is effectively reaching our audience, which allows us to adjust and leverage the intensity of our social media advertising accordingly.

Next, the “authors” tab has allowed us to provide immediate and targeted feedback to our writers. Rather than trying to tease out how many views each author brought us out of aggregated statistics (as we would have to do with Google Analytics), Parse.ly simply tells us how each one is performing in real time. Using this, we have been able to reward the authors who wrote the most successful articles each month and spur our newer writers towards actionable and achievable goals. Whereas before, our authors operated in the dark,­ never knowing which of their articles were the most well received, they are now able to receive immediate and quantifiable feedback. Especially in the realm of satire, this feedback is paramount to the long-term success of the publication.

Lastly, Parse.ly has enabled our publication to hone in on who our target audience is. Over the past six months, the most effective articles were written on topics concerning our neighboring college, Pomona, and timely campus politics. While this may seem intuitive in retrospect, there were competing visions for the publication about whether to focus on more obscure and less political issues. Parse.ly provided us with the quantitative data we needed to identify our most receptive audience.

Overall, Parse.ly became the compass for The Golden Antlers, providing key insights to help guide us out of the dark in analytics.

Show less

Lehigh University

An Analysis on Tags, Devices and New/Returning Visitors at The Brown and White

Students at Lehigh University’s The Brown and White student newspaper use Parse.ly to analyze their visitors and determine their behaviors.

An analysis on tags, devices and new/returning vistors for The Brown and White
Musa Jamshed

Sports Editor, The Brown and White

For the past year, The Brown and White has been using Parse.ly to track the success of different types of content by diligently labeling each of its published articles with tags. Editors then examined these tags to determine what type of readers the publication was able to attract.

Given The Brown and White’s online-first mentality, editors assumed that page views from new visitors were more valuable than those from returning visitors; they were especially interested in content that attracted new visitors who had the potential to expand their readership. Additionally, because mobile engagement has pushed past desktop engagement — and is only growing — the editors were interested in figuring out how to attract more mobile visitors.

As a result, The Brown and White aimed to answer the following research questions through the lens of tags and discuss the trends that accompany them:

  • What are our most successful and least successful tags?
  • Are more of our views coming from new or returning visitors?
  • Do readers engage with our content more from mobile or desktop?
"While there is good reason to write about topics that will interest our readers, newsrooms must ensure that decisions made based on the data do not compromise journalistic principles. View totals are helpful in figuring out the nature of articles that might be successful but should not solely be used to create a formula for attracting audience attention."
— Musa Jamshed, The Brown and White

After assessing the articles published over the past year, and their associated tags, editors at The Brown and White learned the following information to apply to its editorial content strategy:

  • Don’t assume what topics interest your audience. Just because The Brown and White is a collegiate paper does not mean it should cover all campus-related events; some events simply do not generate reader interest. When in doubt, editors can use more specific tags to see where they might be struggling with their coverage of campus-centric articles.
  • Focus on content that appeals to loyal readers. While new visitors are important, burgeoning online publications like The Brown and White should not focus content decisions on attracting new visitors for two reasons:
    • It is difficult to purposely attract new visitors with specific content choices because it is impossible to tell what yielded the new visitors; they may have been attracted, regardless.
    • There actually seems to be a negative correlation between the percentage of views for a tag that come from new visitors and the overall success of that tag.
  • Successful mobile content provides insight into successful content, in general. For The Brown and White, there is a positive relationship between the mobile success of a tag and the overall success of the tag, overall. Because of this, looking at overall success through the lens of mobile success is likely to be a plausible method of analysis.
  • Profiles work! Profiles are one of the most successful types of articles published by The Brown and White because they appeal to new visitors and do well in mobile environments.

Through these initial insights, The Brown and White is confident that it can improve its readership and maintain the integrity of its content.

Show less

Lehigh University

When and Where? Finding the Ideal Traffic Time and the Influence of Social Media

Students at Lehigh University’s The Brown and White student newspaper attempted to determine the publication’s ideal traffic times by studying social media post times and the resulting traffic trends.

What is our website’s ideal traffic time, and how is it influenced by social media posts?
Danielle DiStefano

Editor-in-Chief, The Brown and White

How are social media posts connected to website traffic? The Brown and White set out to identify peak traffic times using several Parse.ly reporting dashboards to analyze its efforts, with the ultimate goal of understanding how editors can best drive traffic to our website through social platforms.

We can now adapt our social media posting schedule to the observed habits of our readers.

We hypothesized that Facebook and Twitter drove the most site views and that our social posting schedule was not necessarily driving views to the site, but rather that our audience was reacting to the time at which we pushed the bulk of our content to the Web.

To test the theory, we first looked at our Fall 2015 social media (Twitter and Facebook) posting schedule. At the time, the schedule was completely arbitrary. Find the schedule posted below for reference.

  • 11:00 a.m. — Image of front page, link to homepage
  • 12:00 p.m. — News
  • 12:30 p.m. — News
  • 1:00 p.m.— News
  • 1:30 p.m.— Sports
  • 2:00 p.m. — Sports
  • 2:30 p.m.— Sports
  • 3:00 p.m.  — Lifestyle
  • 3:30 p.m.  — Lifestyle
  • 4:00 p.m. — Edit Desk or Column
  • 4:30 p.m.— Editorial

For the purpose of this exploration, we continued to post content at the above-scheduled times, while also posting on Twitter and Facebook at other intervals throughout the week to see how our current schedule drives traffic against other factors.

What We Learned About Desktop Traffic

Fifty-seven percent of our readers look at our site from the desktop; we receive the highest engagement from these readers at 9:00 a.m., even though social media posts are never scheduled that early. It seems as though desktop users are checking the site as a part of their morning routine.

Further, Monday, Wednesday and Sunday are our top traffic days from desktop readers. Thursday, Friday and Saturday have the lowest desktop engagement throughout the week. Most desktop engagement comes from evergreen news stories.

What We Learned About Mobile Traffic

Thirty-seven percent of readers access our content via mobile devices. Spikes in mobile engagement range broadly across a 24-hour period, though most take place in the afternoon — the highest being at 3:00 p.m. Wednesday, Tuesday and Friday (respectively) are our most engaged days by page views. This indicates that our readers are engaging with content on the site on the days we are posting to social.

What We Learned About Top Referrers

Our most significant finding — drawn from Parse.ly dashboard data and our own, anecdotal observations  —  is that Facebook provides the most referrals to the website, with 35 percent of social referrals and approximately 38,000 views. Facebook referrals even outnumber internal referrals and direct clicks.

Google refers 25 percent of our traffic, followed by internal referrals from The Brown and White homepage (20 percent of traffic), and then Twitter with 20 percent of referrals.

The Final Analysis

The journalism industry is changing, and we can’t be an online-first publication without understanding what it takes to be successful in the digital space.

Analytics has transformed how our editorial staff functions. Our processes, which were once purely rooted in tradition and habit, are now refocused on data and science. In just a semester, The Brown and White has gained significant insight into our audience. We know more about what they read, how they read and when they read, and have adjusted our daily operations to drive engagement and ensure we continue to bring “all the Lehigh news first since 1894.”

There’s no question that social media is driving views to our website:

  • Facebook is incredibly important to The Brown and White’s website engagement strategy, so we will continue to adapt our posting structure to appeal to our Facebook audience and drive more views to the site. This includes producing longer-form content.
  • Twitter, on the other hand, is best for breaking news.
  • Google is still driving much of our engagement with the site, meaning we must continue to focus on the search engine optimization of our content.

Based on the results of the study, we can now adapt our social media posting schedule to the observed habits of our readers. Instead of using an arbitrary schedule based on sections of The Brown and White, we can share content on social media based on the times our readers are most likely to see it and click.

Show less

Lehigh University

The Perfect Read: Analyzing the Ideal Story Length Using Tags

To better explore reader engagement, students at Lehigh University’s The Brown and White student newspaper tagged all stories with an average “minute read” tag, then matched those reports to Parse.ly’s average engagement time metric in search of the average engagement time for various story lengths.

Reader Engagement Data Makes the Case for Long-Form Content at Lehigh University
Danielle DiStefano

Editor-in-Chief, The Brown and White

Inspired by Medium’s “min read” feature, The Brown and White staff developed a system called “minute-read” in which an editor assigns a predicted read-length tag to every story.

Initially the goal of this system was two-fold. The first was to alert our audience of an estimated reading time per article under the assumption that bounce rates on content would decrease if a reader knew in advance how long they would need to spend reading.

The second was internal — that we could use Parse.ly to determine if the length of our content was affecting the average engagement time with different sections, topics, formats, etc. Ultimately we focused only on engagement due to the placement of an article’s tags on the bottom of the page. We’ve since implemented a plugin that displays a minute-read estimate at the top of all stories.

The ultimate question we were looking to answer: Are readers spending less time, on average, with long-form content than shorter pieces?

Readers spend about the same amount of time on all pieces of content, relative to the length of the article.

Average Engagement Time Across All Content

On average, The Brown and White’s readers spend between 0.1 and two minutes per article, depending on a variety of factors. Content between one and 10 minutes shows average engagement times of less than one minute, and content 11 minutes or longer garners engagement between one and two minutes.

The bulk of the content produced by The Brown and White is considered between a six- and seven-minute read. And of the 335 posts published in our study’s date range, 55 percent of the stories were between 600 and 749 words.

The distribution within the two groups — short stories and long stories — is varied and possibly more dependent on content subject matter than length. For example, student and faculty profiles  generally perform well related to engagement, regardless of length. Student interest stories also have higher-than-average engagement rates.

What Did We Learn?

At the end of the semester, we returned to our hypothesis that readers spend less time, on average, with long-form content than shorter works.

But the results we explored above tell a different story: Readers spend about the same amount of time on all pieces of content, relative to the length of the article. The slightly higher amount of time spent on short and medium-length articles is negligible when looking at engagement with content overall.

As previously mentioned, the range of average engagement times is small:

  • Short- and medium-length stories (1,000 words or fewer) have average engagement of a minute or less.
  • Long stories (1,000–2,000 words) have engagement between one and two minutes.

Relatively speaking, these aren’t much different. The biggest variables are the format and subject matter, which greatly influence how long our readers will spend with our content. It’s also important to consider that our readers may be conditioned to reading content from the “short” group because more than half of our publication falls into that category.

The most significant conclusion we can make from using the minute-read tags to analyze our engagement from a static perspective is that The Brown and White has a very short window to capture its audience and convey its message before readers jump off the page.

However, it’s also a testament to the idea that long-form content isn’t dead.

Over the last few semesters, the editorial board renewed its focus on investigative pieces that often require higher word counts. We feared people wouldn’t read the full article and prematurely draw conclusions about the work from the headline and a few paragraphs. But the results of this project indicate that while we must be strategic in how we capture our readers, that conclusion is the same across all lengths of content, not just one format. However, there are fewer long posts than short- or medium-length posts, so it is more difficult to draw conclusions about them.

This is greatly helpful in determining how reporters structure articles, what kind of content we produce (e.g. infographics, illustrations, etc.), and what that content is about. We have identified a few other tools and technologies that could help us further delve into this question in the future.

Show less

The Medill School of Journalism at Northwestern University

Using the Parse.ly Playbook on NFL Draft Day

During the NFL Draft in Chicago (May 29 – 30, 2016), The Medill School of Journalism at Northwestern University displayed the Parse.ly dashboard on a large monitor adjacent to the production area and recorded the results.

Using Parse.ly During the NFL Draft
Scott B. Anderson

Assistant Professor & Managing Editor, Medill Content Lab and Medill News Service, Northwestern University

The Experiment

During the NFL Draft in Chicago (May 29 – 30, 2016), The Medill School of Journalism at Northwestern University displayed the Parse.ly dashboard on a large monitor adjacent to the production area in which faculty editors and grad student reporters were working on coverage of the event.

The following is what we learned.

The Results

Northwestern with the Parse.ly Dashboard

Students and faculty editors were organically captivated; they regularly looked up for spikes in traffic or walked over to the screen for updates. Classic moths to the flame.

Social Media Noticeably Impacts Site Traffic

Key Findings:

  • An older story had been getting steady traffic, despite all of the new draft-related content. One editor did some strategic tweeting, which helped to bump the older story off the “top daily view” spot.
  • The resulting top story then became No. 1 on Google News search for the player’s name, and No. 3 on general Google search.
  • The VP of Communications for the NFL tweeted praise for one of the newsroom’s stories to his 75,000 followers and traffic spiked noticeably!

Show less

University of Missouri

A New Age in Reporting: How the Columbia Missourian Used Parse.ly to Revamp Reporting

When the students at the Missourian began using Parse.ly, they were able to better understand their audience by tracking referrals, comparing mobile and desktop traffic, and seeing the timing of spikes.

How the Columbia Missourian Made Weekly Reports Actionable with Parse.ly
Elizabeth Stephens

News Editor, Columbia Missourian

The community outreach team at the Columbia Missourian, a daily newspaper affiliated with the Missouri School of Journalism, has relied on the same weekly analytics report to help improve its editorial strategy for the past three years.

When the students at the Missourian began using Parse.ly, they were able to better understand their audience by tracking referrals, comparing mobile and desktop traffic, and seeing the timing of spikes. Armed with these new insights, student Daniela Vidal decided to revamp the weekly analytics report.

Now, they can answer pressing questions: What did we learn about our audience? What does this mean for reporters and editors, the outreach team, the copy desk?

Vidal questioned: “How do we define success? And do our numbers reflect our definition of success? More importantly, does our newsroom care about this?” when she completed the project, the Missourian’s weekly analytics report was more user-friendly and easier to read.

“When I started the process of restructuring the report, I learned that it’s not just sending a report that matters,” Vidal said. “It’s having a newsroom culture that will appreciate and act on that insight. As I talked with editors and teammates, I realized people want to know about analytics quickly, and want an opportunity to dive deeper if they’re really interested.”

Parse.ly’s dashboard helps the Missourian team to better communicate what its readers are doing on the site, and why. Now, they can answer pressing questions: What did we learn about our audience? What does this mean for reporters and editors, the outreach team, the copy desk?

Here are the types of insights that the Missourian has been able to gain from using Parse.ly to help with its new reporting structure:

  • Readers will consume content that’s a little out of the ordinary, and has a visual element that brings the story to life.
  • Our readers value explainer pieces on an issue they might not have known existed.
  • Our audience wants updates and a one-stop shop to peruse what happened and how things are playing out now.

Each week the Missourian team looks at the top (and bottom) 50 stories and works together to determine an average and a median for key metrics. These numbers provide an additional reference point for reporters looking to create content that resonates.

And finally, the team displays the Parse.ly dashboard on a big screen in the newsroom, highlighting the previous day’s top stories and social posts.

Show less