In a media environment where better content is just one click away, publishers should focus intensely on the subject areas about which their communities are most passionate and then use increasingly sophisticated data measures to drive how to distribute this content as widely as possible.
These organizing concepts have guided the approach to reinventing content used by Clark Gilbert and his leadership team at The Deseret News and Deseret Digital Media in Salt Lake City. They also formed the basis for the latest American Press Institute workshop, titled “Transformative Content Strategies,” held Feb. 28 to March 1 in Washington D.C. The workshop was part of API’s Transformation Tour.
While some have wondered whether Gilbert’s strategy has lessons for other media companies, Gilbert and his team are fervent that the concepts adapt universally to any media company, even the most local. Deseret Media, which owned by the Mormon Church, has built its content strategy around six issues that are most relevant to its audience — values in media, faith in community, excellence in education, strengthening the family, caring for the poor and financial responsibility.
That strategy is simple if highly disciplined. First, identity the content your community cares about most. Strengthen your conventional coverage. Transform it so that it is uniquely digital—not just posted on the web first. Involve new forms of content on these subjects, from community contributors to user-generated content and community conversation. Distribute it into social media channels and other spaces as much as possible, listening carefully to the conversation on your key brand areas of coverage.
Then use metrics to determine success and constantly reshape tactics.
This reliance on metrics includes using a dashboard of measurements — which the American Press Institute with the Poynter Institute have made available online as part of a self-directed course, to benchmark whether a company is succeeding. The dashboard includes a few dozen metrics for social media engagement, referral traffic and site design, amongst other categories. API and Poynter also have created an online, self-directed course.
The foundation for the workshop was Gilbert’s theory of dual transformation, which he derived from his work at Harvard Business School with Clayton Christensen, author of “The Innovator’s Dilemma.” Dual Transformation translates roughly into the idea that as you transform a legacy media company, you must also construct a different company that is wholly digital, builds off the existing brand, yet is distinct in culture and personnel. At Deseret, Gilbert has restructured his company into two separate organizations – one focused entirely on the legacy product (which he calls Transformation A) and the other a digital start up (he calls Transformation B). The second company is largely run by people with what Gilbert calls “digital DNA,” many of whom come from outside of the news industry. Most of the growth in your media company, Gilbert warns, will be in the second transformation.
The key concepts inside Gilbert’s strategy for building successful content—online and in legacy media—affect what goes on inside both companies:
Zero in on your desired audience’s passions and plan coverage around those passions
News organization can no longer cover everything. Smaller staffs dictate a new strategy: news companies must choose core topics around which they will concentrate their enterprise coverage and build their brand. In part, this is the nature of the web.
The Internet rewards those who do a few things well, which is different from the traditional newspaper model of covering a little about everything. Today, Gilbert argues, news companies must hone in on topics that matter to the audience they want to have and that they can master at a world-class level.
The process of identifying those interests is complex and involves deep research. But every community is defined by different issues or concerns that matter to citizens.
Deseret’s new digital publishing model.
For Deseret, after a two-year process, leadership arrived at six core issues. They are concerns that matter to people in Salt Lake. They are also issues that have broader audiences nationally and globally—and thus work well on the web. But the idea that this process is unique because Salt Lake is a heavily Mormon community is wrong, Gilbert believes.
The core competencies Deseret built around are: values in media, faith in community, excellence in education, strengthening the family, caring for the poor and financial responsibility.
Nathan Gwilliam, director of social media and vertical channels at Deseret Digital Media, said there is a basic process that companies can go through to uncover the passions that should form the crux of their vertical channels and brand identification.
First, identify passions of potential users. Gwilliam said this could be achieved with focus groups, asking those surveyed to prioritize various factors – such as a specific sports team or industry – by their level of passion. News operations could also have their journalists tap into the community to do qualitative research on what citizens are concerned about. Then deepen your traditional coverage in these areas and create web-only content that takes advantage of digital technology. Next, Gwilliam said, create a growing list of digital channels around these passions. The channels could be Twitter accounts, Facebook pages, a Tumblr, email lists and more. Deseret has dozens of channels in different languages.
The old model for media was to “broadcast one stream of content to everyone.”
The new model is many streams of content broadcasted in a multitude of ways.
That content at Deseret also includes the work of traditional journalists, other news operations, citizen contributors, user conversation, links to bloggers and other social conveners and more. This can be a difficult shift for some traditional news people. It’s one reason this digital work is produced in a different company.
Users often aren’t getting to your site from the home page anymore — adjust accordingly
The home page is just one starting point for your content – and over time a less popular one. Increasingly, users are coming in through the side door as they learn about specific articles and photographs from other sites or by specific referrals from members of their networks.
Gilbert cited as examples Atlantic Media which is now seeing 80 percent of the traffic to its products coming from the side door, and Forbes, whose referral traffic is about 60 percent.
Based on its benchmarking of successful digital publishers, Deseret’s team suggests 20 to 25 percent of referral traffic to a site should come from search engines, an amount that has stayed fairly consistent over time.
About 10 percent of referral traffic, then, should come from link partnerships and at least 15 percent from social media referrals. Chris Lee, president of Deseret Digital Media, said link partnerships, in which someone with an affinity for your content links to it, are not dissimilar to deals hotels enter into with newspapers in order to make copies of the print newspaper available to hotel guests. In either example, it’s a new channel for distribution, potentially getting the product in front of new people.
Facebook is still the highest driver of social media traffic, Lee said. So far, Twitter supplies relatively small amounts of referrals to the core product; Pinterest is rapidly growing as a traffic-boosting vehicle.
With story pages acting as de facto home pages to a large percentage of the audience, it means greater care needs to be taken to present a professional, desirable look on these pages that will encourage further engagement with the brand.
Less is more when it comes to the homepage
The classic web 1.0 idea was about plenty. A typical home page might have had 100 stories or more, grouped in topic sections, in the belief choice would drive more traffic.
That concept of plenty is being radically rethought.
Through close study of audience consumption patterns, managers at Deseret determined they were carrying far too many stories on their homepage in a given day. The optimal number of stories on the homepage at any given time was closer to 20, and the optimal number over the course of a day was 40. The publisher was plugging in several times this amount, overcomplicating the site. When it changed, Deseret experienced a 40 percent jump in page views with the less-cluttered design.
Now Deseret uses a story selection rubric to whittle down possible content to 40 featured items on the homepage daily. But it’s more complex and nuanced than filling the page with stories that are all going to be hyper popular. To the contrary, the goal is to create a mix of stories, some very popular, some less so, to attract a wider potential audience. That breadth among scarcity is the key.
Lee uses a baseball analogy to explain that not all stories are created equal. For each piece of content his staff predicts whether it will be a single, double, triple or home run. A single will likely only get 5,000 to 10,000 views yet there’s considerable less attention and energy devoted to producing it. At the other end of the spectrum is a home run, a story that is expected to get 25,000 or more views and is afforded more care to create. The goal is to have a blend of all categories.
Deseret’s rubric also dictates certain topics cannot be overpopulated. There is a limit on the number of sports stories, for instance, dictates that core brand topics be present and a guarantee that certain key spots on the page be reserved for major enterprise pieces, no matter the expectation of traffic.
The company also knows, algorithmically, what the average traffic should be for any spot on the page at any given time of day and can monitor how a story is faring against the average. This is a way to avoid the trap of wanting all stories on the page to perform equally, all types of stories or all topics.
The life of an article is about 48 hours on the Web
By ascertaining when users were coming to Deseret’s website and reading content, Lee and his team also learned something about how long to keep content posted before it becomes stale and unread.
They found, for instance, that for articles published online after 7 pm – except in cases of urgent breaking news – consumption spiked the next day, not that evening. For content posted between 11 am and 3 pm, interest lasted for one to three hours that day but then picked up the next day.
With the data, Lee said he is able to do a better job of publishing at times when interest in the site is highest. Knowing that, after 48 hours content only gets a 1 percent lift in page views, is useful when figuring out when to switch out stories, he added.
Web-specific content needs to account for at least half of the material on your site
Content online shouldn’t just be a copy of materials published in print. Taking that approach misses out on the innate advantages of the web. By studying other companies that are successful in digital, Deseret has benchmarked that at least 50 percent of the content should be new and distinct from what’s in print.
Gilbert points to several new sources of content to make up this allotment: user-generated content (comment boards, polls), syndication, curation, evergreen stories (polls, lists, quizzes) and community contributors.
A properly-managed community of contributors enhances, not undermines, what you do
A wall mural at Deseret Digital Media emphasizes the right way to think about online community building.
The biggest value of community contributors is tapping into an expertise that you don’t have, yet someone else does. When you create this type of a network the key is disclosing and being transparent about who the contributors are and vetting them as experts. Deseret doesn’t just put out a call for contributors generally but keeps the focus tight when making a request so that the contributors are adding content in topics that are part of the larger strategy for the site and audience.
Once a contributor is selected to begin writing content he or she has to move up a trust curve with a handful of steps along the way to demonstrate the quality of their work. Eventually, once the contributor becomes a trusted voice, they are given greater exposure on the site.
The other advantage of contributors, Deseret has found, is that they are natural social marketers, sharing their content with their networks, which builds up more traffic and recognition for the news organization.
When Deseret leaders surveyed their contributors about why they wanted to write for the site, their three highest reasons were: to publish their work with a professional news organization, fulfill their love of writing and write about a specific topic.
Less important were qualities like financial compensation or recognition. Similarly, the contributors said their top three reasons for continuing to write were: to have an impact, improve their skills and receive feedback on their work.
Readers crave a trusted source to make sense of the wealth of information on the Web — well-executed curation can achieve this
Readers want news organizations to help them navigate the web. This means news site must engage in aggregation of other’s content.
If news organizations don’t take on this task, someone else will, and that is a loss of perceived importance and relevance by a digital audience.
“Audiences will find their interests on or off your site with our without your help,” Lee said. “Why not help?”
In the process, Lee said it helps your writers become even more knowledgeable about their subject matter as they’re directing users to great content in that subject generated by other experts.
Curation can be a delicate matter, however. Deseret has consulted Columbia Journalism School Associate Dean Bill Grueskin’s work on ethical curation and aggregation to set up procedures so they’re not stealing an article from another organization but transforming it into a new product. At Columbia, Grueskin has modernized the curriculum for graduate students by supplementing traditional journalism with new content practices such as curation.
Use analytics to measure and guide
The other underlying principle of Gilbert’s approach is that analytics takes the guesswork out of understanding reader behavior and site performance.
News organizations have traditionally made editorial decisions by intuition, using gut feelings to predict which stories will perform well and where to physically place them on the page. That strategy won’t work in a competitive media environment, Gilbert argues, where digital competitors are providing content and measuring it relentlessly. Digital platforms carry the potential to bring in much more sophisticated data sets and analytics to all parts of the decision making process.
Using numbers to gauge effectiveness of content, however, is no small step and is actually part of a larger cultural shift for news companies. Editors are not accustomed to being held accountable for their decisions about content based on numerical standards. Today, that is unavoidable. The key, Gilbert’s team argues, is to have a concept of what makes up your brand and not simply build expectations around what generates the most click views. Developing the numerical goals for content involves critical decisions about values.
Lee recommended a range of analytics tools for news organizations, some suited for instant feedback on individual pieces of content and others better for bigger-picture thinking.
Deseret uses Google Analytics to assess “who goes where,” Lee said – where users enter the site, the path they take between pages and at what point users drop off from their product.
Among the real-time sources of data Deseret uses are Visual Revenue and Chartbeat. The team uses Chartbeat for analyzing social sharing of content and looking for patterns around which subject areas generate the most shares on social media.
It uses Visual Revenue in “content merchandising,” deciding which items should be where at specific points in the day. The software product, Lee said, provides an average click-through rate for every position on your site over time. Using that data as a barometer, Visual Revenue can compare a current piece of content to the average to determine if the story is performing adequately in that specific moment. If it’s not, Lee said, it’s easy to adjust. The software also will suggest a more reader-friendly headline and then test whether the suggested headline fares better than the original headline.
The results are real. Traffic has surged. The company is also deriving more than 40 percent of its revenue from digital.