Offered in both English and French, the JRP Canada team has created a slideshow template to support academics, students and speakers in presenting accessible presentations to their audiences. The visual and oratory tips included that can be adapted and applied to any presentation and media formats.
The increasing integration of automation into journalistic life is often coupled with discussions of job threats, loss of identity, and an increase of disinformation.
And while fears are not unfounded–in every field, artificial intelligence comes with a slew of ethical and technical challenges–industry experts believe that the evolution of tech can change journalistic roles and work for the better.
In a panel presented virtually at Toronto Metropolitan University’s Creative School and moderated by School of Journalism professor Nicole Blanchett, digital journalism specialists highlighted that despite an onslaught of discourse suggesting otherwise, AI isn’t replacing journalists – it’s letting them do more of the work they were intended to do.
“We all became journalists to change the world, to find and tell news that can change people’s lives,” said Sonali Verma, the deputy head of audience at the Globe and Mail and the director of business development of the Globe’s Sophi.io, an artificial intelligence and machine learning platform. “Not to move headlines up and down the page based on how many clicks they’re getting, or how many subscriptions they’re generating.”
AI supports newsrooms, she notes, by taking care of “grunt work” and re-orienting human journalistic resources towards more rigorous reporting. At the Associated Press, for example, automation has freed up 20 per cent of its reporters’ time that would otherwise be dedicated to corporate earnings coverage. The Canadian Press is also using automation in order to dedicate more time to enterprise reporting.
Lucas Timmons, a news automation developer for Torstar, said the ascent of artificial intelligence means the automation of time-consuming reporting, particularly in the sports and business domains of journalism, where programs can be coded to quickly translate scores and stats into news stories for audiences.
AI has also led to the emergence of new journalistic jobs like Timmons’, which sees him writing software that finds, cleans, and processes structured and unstructured data into written narrative stories. And that of Katie Kutsko, an education and strategy manager for the American Press Institute, and another of the event’s panelists, who trains newsrooms on how to strategically use audience data.
Referring to making audience data easily digestible and actionable for journalists, Prasanna Rajagopalan, the director of journalism and programming at CBC Toronto, said, “We really want to ensure that our journalists are focused on the storytelling and getting to the heart of the matter.” And with journalists being expected to cover multiple beats, perform a number of journalistic roles, all while consuming and aggregating a massive repository of data into stories, AI, says Rajagopalan, simplifies the content-creation of an ever-complex and ever-changing industry.
Now more than ever, the panelists agreed, no matter the role, being a journalist means working and understanding data. And while that’s bolstered a reporter’s job description, it also makes for more dynamic roles. According to Verma, adopting “radical data transparency” in the Globe and Mail has created a newsroom culture where every journalist is given the tools to look into and question the data themselves, embracing their journalistic disposition as “professional skeptics.”
Journalism is evolving and it’s becoming more intertwined with tech, so if you’re technologically or mathematically-inclined, says Timmons, “this is a growth area” and J-schools have to be training their students to work with this machinery.
But in every facet of AI, journalists still remain indispensable to the journalistic process. Currently, no evidence suggests that automation has led to a loss of employment. The true danger lies in allowing machines to exist independent of human supervision and editorial decision-making.
No matter what tool you’re using, says Kutsko, whether it’s Sophi, Google Analytics, or Metrics for News, newsrooms are invariably encouraged to “look at data as a signal,” said Kutsko
“It’s never going to replace your human editorial judgement – it’s going to enhance it.”
Is there a difference between the ideals that journalists believe in and how they practice their journalism? And, if so, what does that mean for media workers and news reporting? That’s what a team of researchers from 47 countries around the world, including Canada, is exploring with the Journalistic Role Performance project.
With the idea of what journalism is and what counts as journalistic work being challenged everywhere, the international project lead believes the timing of the study is crucial.
“Understanding how and why journalistic roles are performed around the world is fundamental for this moment in journalism, in which the profession, its practices, and its credibility are increasingly being questioned,” said Claudia Mellado, a professor at the Pontificia Universidad Católica de Valparaíso in Chile.
The JRP project goes beyond the idea of industry defined roles such as covering a particular beat and examines journalistic role performance using three categories: the presence of a journalist’s voice in their story; the relationship between the journalist and news outlet to those in power, economically or politically; and the journalist’s relationship with the audience — are they seen as citizens to inform and support or customers to sell to?
For analysis, journalistic roles are divided into six dimensions. The first is the interventionist role, where a journalist takes sides or promotes action as part of a news story. The second and third roles are about power relations — the watchdog’s role includes critiquing the government, while the loyal facilitator supports government narratives. The relationship with the audience is the focus of the last three: in the service role journalists cover everything from consumer tips to health care recommendations; in the infotainment role, they create content that is geared more to entertain than inform; and in the civic role, they put the perspective and rights of citizens at the forefront of coverage.
This is the “second wave” of the JRP project. The first ran from 2013 to 2018, with researchers from 18 countries focusing on newspaper journalism. A recently released book highlighting findings from the first wave, Beyond Journalistic Norms, is edited by Mellado, and works to explain differences between “normative visions and actual practices.”
Sites of study in Canada include CBC.ca, huffingtonpost.ca, LaPresse, the Globe and Mail, the Toronto Star, the National Post and broadcasts from Radio Canada, CTV and Global. Throughout 2020, on 14 different days, researchers collected thousands of stories produced by these news outlets, and thousands more were collected from other news outlets by researchers from around the world. All of the stories are being analyzed using a codebook shared by the global team, allowing for a cross-country comparison of the similarities and differences in the prevalence of journalistic roles in news content on different media platforms. Additionally, international research teams are tracking how many stories being studied are related to COVID-19.
An anonymous survey is also being sent to hundreds of journalists across Canada, and thousands around the world, who created the content being studied. The survey asks questions specific to theoretical constructs of role performance, but also on practical things like the use of metrics and analytics in newsrooms and how much autonomy journalists have in selecting the stories they develop. The survey results will be compared to the content to help determine gaps in ideals versus practice.
For example, if you ask most journalists, they’ll tell you a critical part of a journalist’s job is holding those in power to account. But despite the level of importance ascribed to that function, devastating cuts in newsrooms mean there are fewer reporters who have the time or resources to perform it. And, changes in practice are leading to new priorities and definitions of what encompasses journalism.
An audience engagement editor participating in the Canadian study insisted they’d never been a journalist, but rather a “media person.” The reason? Even though they created content for their organization’s website and had written and worked as an editor for a number of news organizations, they were more interested in “the reader” than “the story.”
A Canadian journalist talked about the lack of resources to do “important work” and demands to create more listicles or “10 things you can do”-style stories because editors felt they were popular.
Another noted, “You can spend so much time writing an investigative piece that gets one share and the kitten story gets hundreds.”
A freelance journalist said news is now “what people want to read.”
What journalists believe to be ideal practice and the work they actually do does not always align, and the gap in that alignment is the focus of the JRP project.
“One of the main contributions of our project is that it shifts the focus from the idea of professional roles as taken-for-granted behaviours, to that of professional roles as socially constructed, and comprised of ideals and also practices,” said Mellado.
Traditionally, journalistic practice has been compartmentalized. But many surveyed journalists from the first wave of JRP didn’t fall into traditional paradigms, and the importance they placed on specific roles had little relation to whether they performed these roles.
For example, the infotainment role, creating content with an emphasis on entertaining the audience, had a more predominant global presence than the service role in content that was analysed, and infotainment wasn’t siloed to a particular type of newspaper or beat, but spread across topics and cultural contexts.
The first wave found that journalistic roles are becoming hybrid in an industry where journalists are expected to cover more ground, often working as freelancers.
“They work for different media outlets, they produce stories in different beats, and write for different audiences all at the same time, so it makes sense to think that they are constantly negotiating their ideals with situations that require the combination of specific performances,” said Mellado.
“Sometimes, they are watchdogs, other times they are entertainers, sometimes they intend to perform a service function, and others they do everything together.”
Although roles are changing here in Canada, there is not as much polarization of media as seen in some countries, particularly the United States, nor the same level of restrictions as seen in media systems with governments that have more control over the circulation of information.
There is also broad agreement on the importance of journalistic ideals such as verification and working to share evidence-based information
“The journalistic field has been very normative in nature, dictating to us whether journalism can be considered ‘good’ or ‘bad,’” said Mellado. “And in my opinion, the excess of normativism is not helping us in our understanding of how journalism continually redefines its rules in different logics and historical times. Norms also evolve.”
Journalistic practices are complex, as explored in research on the use of metrics and analytics in newsrooms. As a result, Mellado emphasized that journalistic roles should not be studied out of context: “Journalistic practice is embedded in organizational and industrial contexts shaped by social factors: economic, political, cultural. This idea is common sense, but it is still not sufficiently recognized.”
Although much was learned from the first wave of JRP, researchers recognized that the initial study had significant limitations, including the fact that only newspaper content was being analyzed. That is why the current, second wave is taking a multiplatform sample, analyzing content from television and radio newscasts, websites and newspapers.
Mellado said another issue with the first wave of the study was that only hard-news stories were examined, blocking the understanding of “how professional roles may play out differently in other topics.”
This time, all types of news stories are being analyzed, from politics to lifestyle to entertainment, and, as a result, a wider range of journalists surveyed.
Finally, Mellado identified a limitation with the sample of countries involved in the original study; there were none from Oceania, the Middle East or Africa. Now, there is representation from each of these areas and even “different areas that make a difference within each region or continent.”
There are, however, some unique components to the Canadian study. Its survey includes a section on the working conditions and roles of journalists in the pandemic, and there is a qualitative component including interviews with some of the journalists whose work is being analyzed, along with limited pre-pandemic newsroom observation.
The international dataset being collected by the JRP study is of vital importance in understanding the ever-shifting field of journalism and providing new understanding with regards to the impact, or changing impact, of differing media systems on journalistic practice. In another publication highlighting the first wave’s findings, researchers identified that, now, “practice can differ more clearly between journalists and media organizations than between countries.”
From a Canadian perspective, the JRP project will provide critical insight into journalistic practice in Canada in the unprecedented period of the pandemic. Along with the material required to perform analysis for the JRP project, Canadian researchers have also collected supplementary data from all Canadian sites of study from January to June of 2020, documenting corrections and editor’s notes; the ratio of original to third-party content being published or broadcast; and temporal data related to the number of days well-performing stories are kept on website homepages, creating a rich dataset of journalistic output.
Through the JRP study, Canadian researchers hope to gain greater knowledge of the changing roles of journalists and content creators with a goal of better understanding what skills should be taught in journalism schools.
By combining the study of news production with research on professional roles, findings will allow both industry professionals and educators the opportunity to critically reflect on theoretical narratives of practice versus actual performance, in order to better examine how journalists might best serve the public.
“Norms are good for constructing discourses of occupational legitimacy,” said Mellado, “but we shouldn’t think about journalism in the abstract, and we should not make absolute claims about the relationship between the practice of journalism and concepts that are quite culturally bound, such as democracy, objectivity or autonomy.”
The future of digital storytelling is intrinsically tied to the use of audience data, and that’s leading to fragmentation of the audience.
I’ve been studying the use of metrics and analytics for years, but my recent conversations with journalists were part of a broader study called the Journalistic Role Performance (JRP) project. In cooperation with 36 other countries from the Global North and South, we’re examining the potential gap between journalistic ideals and practice through a content analysis of tens of thousands of stories from multiple platforms and the survey of thousands of journalists.
Here in Canada our sites of study are The Globe and Mail, Toronto Star, and National Post; CBC, CTV and Global’s national television newscasts; the CBC news website and the now shuttered HuffPost Canada; The World at Six broadcast on CBC radio; and our French sites of study, LaPresse, TVA’s Nouvelles, and Radio Canada’s l’heure du monde.
No matter what the organization, there is an expectation of digital presence. Journalists don’t just work in television, print, or broadcast, they are contributing to multiple platforms, often with a primary focus on digital delivery. And the data accessible as a result of digital delivery are changing what happens in newsrooms.
There are a variety of metrics that influence story development, from traffic measures like pageviews that give organizations a sense of the broader interest in a story–and continue to be tied to advertising mechanisms such as page impressions–and metrics like time spent that speak more to engagement, or how people are interacting with the content that is produced.
Increasingly, news organizations are gearing content to their loyal readers, versus relying on a broader audience as a primary source of revenue, and often content created for those loyal readers is put behind a paywall. This means different segments of an audience have different access to a particular organization’s stories–and different content might be promoted to different readers, depending on past behaviour, what an algorithm determines might be of interest to them, or how the content might make money.
This shift with regards to digital revenue means many organizations are focusing on subscriptions versus fly-by readers. That means loyal readers are deemed more important and metrics like conversions, or identifying what content someone interacted with before becoming a subscriber, can be more significant than pageviews–although pageviews continue to be a significant marker for ad sales and an easily understood signifier of success.
In a conversation with analytics experts from the American Press Institute, it was noted that media organizations are “segmenting” audiences and paying attention to multiple metrics for different reasons, even within what would have previously been thought of as one audience. It has a tool that can create blended engagement scores based on different goals for these audience segments.
For some articles targeted to get wider traffic, pageviews might be more important, for content created for loyal readers time spent and conversion might be more important. Organizations can place more or less weight on specific metrics for specific audience segments, in relation to how each type of audience might build revenue and/or as measures of success.
At one of our sites of study, the Toronto Star, most of the content is paywalled, but there are stories that include what they describe as affiliate links that appear to be open access. If a reader follows through to a product link that is embedded in one of these articles and buys that product, the Star receives revenue.
A study participant noted these articles were generally lifestyle stories written by freelancers, but also noted that it could be hard for readers to discern they’re different from any other story. Unrelated to the JRP study, Toronto Star columnist Vinay Menon wrote, “I always chuckle and then curse my overlords, the Gods of Fluff, when someone gets snippy about a column I’ve written and asks, ‘How is this news?’ It’s not news! Who said it was news?”
There’s research to support that readers can have trouble differentiating between paid, sponsored, or even ideologically-driven content over actual news, and that even the use of in-story descriptions to help clarify a story’s intent (which the Star does), has limited impact on perceived news credibility.
When something “looks like” news to a reader/viewer, it will often be perceived as news, a byproduct of media logic. The definitions journalists and scholars apply to journalism don’t necessarily hold meaning with audiences.
The now shuttered HuffPost Canada also demonstrated how processes that used to be siloed within the marketing or advertising world are crossing over into digital newsrooms. It developed a system where writers had to pick who they were writing for before starting a story, with profiles of imagined readers created using audience data: Adela for “young millennials”; Adam for “middle millennials”; and Diana for 50-60 year old women.
One participant described it as “classic magazine marketing” where you pick an “audience funnel.” The practice was designed to reinforce that no story was targeted to “everyone” and that they needed to “place their bets” on who was most likely to read it. When a story was completed, the audience engagement team had a checklist of what needed to be done in order to ensure the story was as visible as possible, including whether it was posted to Facebook and Twitter and pitched to news curators such as Apple, who would push the story using their own algorithms. I’ve seen similar processes to amplify content used in other newsrooms in Canada, England, and Norway.
Although social media are an integral part of story distribution at many outlets, and reporters are often expected or at least encouraged to develop their own brand and promote their own work, this can be extraordinarily time consuming and open up reporters to toxic online abuse, particularly those who are racialized, LGBTQ+, and/or women.
In order to survive, newsrooms have to find new sources of revenue and explore new ways to tell stories. Audience data can help inform that process–and that’s not always a bad thing. It can lead to innovative storytelling and greater participation and input from the audience on where newsrooms should be spending time starting at the story development phase.
But, as noted by researchers at the Reuters Institute of Journalism, it’s also leading to people who might most benefit from good information having the least access to it, because they can’t or don’t want to pay for what’s behind a paywall, or news organizations see no fiscal benefit in trying to reach them. The best news is often developed for audiences that are already well-served, and white.
Additionally, although all news sites have to explore different revenue mechanisms, having what might be considered the most informative content paywalled, while content that is less impactful is fully accessible raises some critical questions with regards to journalistic ethics and the overarching goal of journalism to inform the public.
Developing the best digital stories, developing the best mechanisms to make money from digital stories, and having the best ethical practice in relation to sharing content don’t always align–but take out the word digital and you can pretty much say the same for all news from an historical perspective. The big difference now is the extensive and direct impact of data on the way stories are shaped, promoted, and shared.
From a bottom-up editorial process to prioritizing diversity and using traditional marketing practices to develop journalistic stories, HuffPost Canada was a digital-first innovator. Then it was shuttered.
My interviews and observations at HuffPost were part of data gathering for the Journalistic Role Performance project, an international effort between 37 countries exploring if there’s a gap in journalistic ideals compared to practice.
After collecting thousands of stories in 2020, then coding them and surveying journalists from the news organizations who produced those stories, we’re now getting to the analysis stage. And, by coincidence, we’ve captured some of the last days of HuffPost Canada.
Right before the first COVID-19 lockdown, I observed practice in the HuffPost newsroom. With wood-planked floors, high ceilings, exposed brick and lots of natural light, it had a different look and feel to it than many legacy news organizations. The “feel” reminded me a lot of Citytv Toronto when it was located on Queen Street West, where I worked as a television news producer and writer in the 90s and early 2000s. One reason was the young and diverse staff. CityPulse News, as it was known back then, was the first to make an effort to reflect the community it served.
In an interview done for students in my first-year journalism class, CBC anchor Dwight Drummond reminded me that when we worked at City together the motto of the newsroom was “everybody welcome, everybody belongs.” He said, for the first time, being a person of colour who grew up in public housing was seen as an asset.
Newsroom diversity industry-wide may have improved somewhat in a few decades, but there’s still a long way to go based on a report from the Canadian Association of Journalists that found “almost half of all Canadian newsrooms exclusively employ white journalists,” about 9/10 have no Latin, Middle Eastern or mixed race journalists, 8/10 no Black or Indigenous journalists and 2/3 have no Asian journalists.
At both HuffPost and City, the focus on diversity didn’t stop with the person reporting or cutting the stories, but flowed through to the use of sources and experts. One HuffPost editor said,
“Our big thing is that we normalize diversity. We don’t have special sections, we just do it–and if that approach can influence other media, that’s a marker of success for us.”
In terms of newsroom hierarchy, there was a “striking difference” in practice at HuffPost, according to one reporter who formerly worked for a legacy newspaper. At her previous job, the editorial process was completely top-down. Decisions on what she covered were based on what editors “felt” should get published. At HuffPost she came up with her own ideas. Another employee said everyone “was encouraged to contribute and innovate, and even dissent if they felt a story was inappropriate.”
At an editorial meeting I attended, there was a lot of talk about what was trending, how search engine terms could be leveraged to attract readers, and how social media could be used to promote both stories and engagement with particular segments of the audience. However, there was also a lot of discussion about politics, including an investigative piece coming out of Ottawa.
A study participant from another news outlet acknowledged HuffPost Canada did some good work, but questioned why it was part of our research. He said they were “national” only because anybody could “click on them” but their “repertorial footprint” was “pretty thin.”
This was acknowledged by HuffPost. One editor said it was a small team and there was “no illusion” that they could cover everything. They relied on agencies like the Canadian Press for stories they didn’t have the resources for, and encouraged reporters to focus on what they were passionate about and develop stories they’d be “remembered for.”
He stressed they weren’t just going for “cheap clicks” and that speaking “truth to power” and giving a “voice to the voiceless” was their “brand.” However, he also said there was “no shame” in doing viral stories, and didn’t understand why they were somehow considered “dirty” or labeled as “clickbait.” He compared them to the “water skiing squirrel” of TV news–I wrote about that squirrel(s?) more than once at City. Ratings mattered then and now–there were overall traffic targets at HuffPost.
Early analysis from our JRP data shows some notable differences in HuffPost Canada stories compared to the other news organizations in our study, including more prevalence of the use of first-person and the journalist’s point of view. Although recent research shows that the majority of news consumers might prefer a more traditional, impartial style, younger and/or left of centre audiences are more open to it, and there’s evidence that’s who HuffPost Canada was targeting.
Based on audience data, it had created imagined readers including Adam, a “middle millennial” who had a partner named Taylor, and Adela, a “young millennial” who was on Insta at 10 p.m. Reporters were supposed to use this marketing-style audience funnel to put a “face” to who they were writing for before starting a story.
HuffPost recognized that topics like parenting that may be considered as “lighter” or less “important” than politics by other news outlets were really important to their readers. An editor said their role was to ensure all information was delivered in a way that helped people “make smart decisions” about their everyday lives. There were two important questions: “How does this affect me and why should I care?” We called that news you can use when I was at City.
She said HuffPost’s “digital native stance” was based on acting as a “great equalizer” in the distribution of information, because information wasn’t “only for people who can afford newspaper subscriptions or maybe have reached a certain level of education or reading comprehension.” Serving only the most educated and affluent news consumers and the use of paywalls have been noted as growing concerns by the Reuters Institute of Journalism.
Understanding how information moved digitally, following a proprietary process of search engine optimization, and amplifying content through social media were part of everyone’s job, as was building community. On a Facebook page with a focus on housing, for example, information was shared no matter where it came from, including other news organizations. Additionally, they always responded to suggested corrections from readers to try and show a “human face.”
An eye to engagement was also embedded into story-packaging practices. One editor noted that instead of labeling something “politics” it might be called “pointing fingers” in an effort to boost interest. The added pressure of having to package her stories and think about SEO was noted by the HuffPost reporter mentioned previously. When she said she was still learning but getting better at it, from another desk someone shouted out encouragingly, “You are!”
That sense of camaraderie was visible on multiple occasions. As another editor noted, “If someone has a larger project they’re working on, others will pitch in to help” –it was part of managing limited resources, but also evidence of a tight-knit team.
I’m sure there were downsides to working at HuffPost Canada. I didn’t get to spend enough time there to get the full picture–particularly for those who might have been doing contract or freelance work. And anyone who worked in the City newsroom during the timeframe I was there could offer a laundry list of serious issues. But there were ingrained priorities for both of these newsrooms that are transferable to others: amplifying diverse voices, connecting with the community, and breaking traditional formats and practice in order to engage the audience.
When asked to describe the impact of the closure of HuffPost Canada, one study participant emailed this response:
“We combined relevance with irreverence, having fun with the news when appropriate, and digging in with our considerable editorial talents on investigations whenever possible. We prioritized diverse communities’ perspectives and sought out — and featured — the voices not often heard from, and Canadians are seeing less of that without HuffPost Canada’s contributions to the landscape. That feels like the greatest loss, and hopefully as our journalists and editors get snapped up by other outlets, is a change that’s soon seen elsewhere.”
I hope so, too.
An edited version of this article was previously published on The Conversation Canada.