Often, when presented with an analytics platform people focus on metrics. What are the numbers telling us? How many people are reading my post right now?

Last month, at the News:Rewired conference, hosted by Journalism.co.uk, we asked 40 journalists, editors, social media editors and audience development to not think about numbers, but to think about their audiences.

What did they want to know?

• Who are they? How old are they?
• What do they like? What stories do they like?
• Do they keep coming back? What do they like that keeps them reading?
• What will they be willing to pay for?
• How much are they willing to consume? Longer pieces? Lots of shorter ones?

Digital publishers of all sizes ask these questions about their readers. Other internal teams, like marketing or analytics, know some of these answers through survey data, demographics from subscriptions, and market research. However, that information often isn’t  shared with the editorial teams, or it’s too broad to help specific sections or writers.

Making it about audience insights, not numbers.

What did we tell the audience at News:Rewired?  Editorial teams, however, know their content inside and out. We combined content and analytics to find insights into the audiences reading their stories.

We didn’t we set up computer screens with excel files and teach anyone how to use pivot tables.  In fact, we stripped away all of the metrics, page views, unique visitors, sessions – all of it, gone. We started with basic information that every editor should be able to get their hands on: the top 25 posts, top 25 tags, and top 25 referral sources for a sample site.

The journalists, working in teams, examined this information to answer questions that digital media companies ask  internally. Each team was tasked with creating an audience profile and a recommendation for a specific role:

• Social Media: What would you tell this site’s social media editor to focus on?
• Editorial Strategy: Do you see an area that we aren’t covering as a vertical that we should be?
• Distribution and Discovery: What can we try as a new distribution strategy?
• Audience Development: What can we do for audience development to attract new readers?

Without numbers or  excel files,  without any data analysts and without even knowing what the website  was, they created an accurate audience profile, including political views, geography, brand preference for technology, content format preferences, gender and interests. All based on the content the readers liked!

Once they had a clear picture of the audience, they could focus on suggestions that would improve the reader experience, encourage more people to visit the site and keep loyal readers on new sections and posts. Here are their ideas:

SOCIAL MEDIA

Some of the top referral  were the twitter accounts of industry-niche celebrities. The social media teams said they would encourage reporters to engage directly with people on Twitter that were already sharing and talking about their stories.

A number of the articles saw a huge percentage of their traffic coming from Reddit and sub-reddits, so they said they would spend resources to educate the whole editorial team on how to use Reddit, more so than they might at another outlet.

A final recommendation was to increase the creation of entry level, evergreen content like tutorials and primers on topics that popular stories covered to attract new audiences and ease them into the articles. Many of the existing articles were very in-depth and helping new readers understand them better could mean they stick around to read more.

EDITORIAL STRATEGY

The group tasked with identifying possible new verticals saw some opportunities based on the most popular products the site covered in their tech-product review section. The readership of this site frequently liked Android product reviews, so they suggested possibly creating a vertical that would be dedicated to both Android products and stories about the type of lifestyle Android-product users lead.

Not all decisions can be based on basic content information, and this group wanted more data. They told us they’d need more information based on the breadth of readership: were readers reading only stories in one vertical? Did they cross sections often? And how many stories would they read? Being specific about type of data needed  is a critical skill when doing analysis.

DISTRIBUTION AND DISCOVERY

Members of the workshop discussing distribution strategy honed in on the combination of powerful Twitter users and suggested that opening up a guest posting area of the site could be a powerful strategy. Based on the obviously strong and focused community that existed on this site, allowing user generated content and curating guest posts could create a large amount of easily shareable content to engage existing readers and find new ones.

AUDIENCE DEVELOPMENT

Thanks to some smart developers at the sample website, not only did the tags the group looked at give an indication of the topics the content covered, they also showed what types of story formats were working: longer form reporting, short summaries, and regular length reporting*. The group that longform pieces did really well, but there were less of them. They recommended assigning more longform pieces.

*This is a great way to use meta-data to make decisions on more than just content. Using Meta-data and Tags can be incredibly powerful!

A Take-home Quiz.

Try this with your stories. This exercise can be done at any digital media organizations within sections or by individual authors to find new opportunities and make sure your editorial team understands their readers. The information can be gathered through most analytics tools, though if you’re a Parse.ly client, pulling top posts, tags and referrers couldn’t be simpler:

Picture from  John C Thompson.