Boundary Work, Interloper Media, And Analytics In Newsrooms

Abstract

Amid the increasing use of web analytics to gage the success, present and future, of news content and related news products, this article focuses on in-depth interviews with the suppliers of those analytics: web analytics companies. Drawing on the concepts of boundary work and interloper media, this article examines how managers for these companies understand and position their work in relation to news production as a boundary object. The larger finding of this article is that while web analytics companies seek to understand and address news production values and norms without assuming responsibility as journalists, they foster profit-oriented norms and values in newsrooms by introducing web analytics as disruptive, connective, and routinized in news production. By offering a product that they (i.e. web analytics companies) need to modify on a continuous basis because of changes in the structure of the web and audience behaviors, they foster a milieu of constant experimentation with old and new products. This and other findings provide a portrait of the development of norms and values in journalism as these companies gain a greater acceptance in newsrooms and increasingly influence news production.

Keywords:

• boundary object
• interloper media
• interviews
• journalism
• managers
• metrics
• newsrooms
• news production
• web analytics

A relatively small number of web analytics companies are now embedded in a sizeable number of global newsrooms, either physically or through the information they provide, each working to ease some of the tension journalists feel with audiences while also working to boost news organizations’ revenue models. The impact on the news process these companies are having has been a major focus of media scholars, who have given voice to journalists and other news producers without much attention focused on the companies articulate their role (see Ferucci and Tandoc 2015; Hanusch 2016; Lewis and Usher 2016).

Amid these developments, and given these companies’ distinct values, norms, and practices to journalism, it is critical to examine how web analytics companies understand and perceive their place in news production (cf. Petre 2015). If these companies increasingly influence news culture, how exactly they position themselves in the news process and how they communicate their norms and values to news organizations is practically and theoretically important. Using in-depth interviews with web analytics companies’ managers working with newsrooms globally, this article examines the perceived roles of journalism web analytics companies as well as the ways in which they maybe altering the boundaries of traditional journalistic values and norms. This study explores how these managers seek to coordinate their interpretation of news production in journalism. As these companies become more vital to news production, it is urgent to understand how web analytics professionals and journalists become entangled in that process.

To unpack these issues, this article is divided in four sections. The first section outlines the concepts of boundary maintenance and interloper media, each contributing to the formation of web analytics managers’ professional logic with news production. The second section highlights the methodological choices made in this research, while the third section explores how these managers coordinate their professional boundaries with those of journalists by discussing the roles that these managers understand and perceive they have in journalism. The fourth section concludes that while these companies do not position themselves as journalists or as key players in a shifting journalism culture that increasingly includes web analytics, they nonetheless see themselves as an emergent, if not already integral, part of the news process. This suggests that web analytics companies, which are profit-driven and external to journalism, are wrestling with issues of self-perception, pitching themselves as necessary to journalism on one hand while not wanting to claim responsibility for any of the changes their involvement may have on journalism culture on the other. In other words, web analytics companies see themselves as improving practice for individual journalists and the news organizations they serve but not as drivers of cultural or institutional change more broadly.

Literature Review

Web analytics and optimization companies such as Omniture, Accrue, WebSideStory began surfacing in the mid-1990s. These mostly for-profit companies focused on collecting and processing data into information, providing key performance indicators, and formulating strategies for web optimization for varieties of businesses and organizations. After acquiring Urchin Software Group in 2005 and launching Google Analytics one year later, Google worked to make such data more accessible to the public (AP 2005). Google Analytics now allows users real-time monitoring of web traffic in more easily digestible formats than some other web analytics companies, though there are data limitations based on page views, organizational sizes, affiliations, and user goals.

The success of Google Analytics, which is now used by nearly 75% of all Fortune 500 companies (Anderson 2016), brought web analytics to the forefront for businesses as they rushed to make their brands, content, and audience engagement appropriate in digital environments. Web analytics companies made their debut in newsrooms shortly after the launch of Google Analytics, with Chartbeat, Newsbeat, Parse.ly, and others providing audience analytics and metrics targeted at improving news delivery and audience engagement. By 2011, these companies worked with dozens of prominent news sites (Sonderman 2011). Today, these companies and social media analytics companies (e.g. Hootsuite, Facebook Insight, the Twitter activity dashboard) constitute an important part of the market of editorial and generic analytics tools that are reshaping how newsrooms operate.

Initially used to help predict which stories, visuals, and advertisements would draw the most audience traffic, news organizations can now use analytics to determine what content should be published, where it should be placed, how long it should remain in one place, and when it should be followed up on (Lee, Lewis, and Powers 2012; Levy 2016). Even though newsrooms across the world have incorporated web analytics in their daily practices, scholars are hesitant to suggest that journalists should adopt audience-driven data into news judgment practices (Anderson 2011a, 2011b; Nguyen 2013; Tandoc 2014; Zamith 2016). Web analytics are reshaping the culture of journalism from one wherein editorial managers and publishers used numbers and complex data as gages of performance that were funneled to journalists to one where they use simplified data and visualizations to drive news content, often including journalists early in the process. Scholars have studied the ways in which web analytics have been adopted as metrics of success for content and audience engagement (Bunce 2015; Duffy, Ling, and Tandoc 2017), how they have forced journalists to rethink their professional processes (Tandoc and Thomas 2015), and how journalists are working within a newsroom culture that places, in some cases, more value on web analytics than journalistic intuition (Hanusch 2016).

Despite the increased presence of web analytics in the newsroom, journalists appear to be at odds with how exactly they fit into news production. Academic studies that have analyzed journalists’ use and perception of web analytics indicate that journalists recognize their utility but do not themselves fully embrace analytics as necessary to successfully produce news (cf. Zamith 2016). This is, at least in part, due to editors’ traditionally tight hold on web analytics as well as the complexity of web analytics, which have not always been presented in ways that are easily understood or put into practice (Petre 2015). Even though journalists can make use of publicly available alternatives (e.g. Google Analytics, Clicky, Open Web Analytics) that maybe more approachable, they are increasingly faced with interpreting web analytics being shared by editors. This uptick in newsrooms’ analytics uses indicates the important role of analytics companies in news production and raises questions about the ways in which these companies position themselves within that process. While journalists may see these companies, or at least the information they supply, as challenging the traditional news production (e.g. placing value on numbers rather than intuition), the ways in which web analytics are enmeshed in the news production suggests that web analytics companies maybe challenging journalistic boundaries and repositioning themselves less as media interlopers and more as key actors in news production.

Maintaining Boundaries Through Objects

Professions engage in a certain degree of boundary maintenance while trying to gain authority over particular definitional domains. They seek to establish “cultural authority” while maintaining symbolic contests to marginalize non-professionals (Abbott 1988; Carlson and Lewis 2015; Gieryn 1983; Lamont and Molnar 2002; Starr 1982). In journalism studies, there are many cases of occupational cultures working together to produce news. For example, Baack (2017) argued that in the case of data journalism and civic technology, tech companies could “interlock between practices of facilitation (enabling others to take actions themselves) and gatekeeping (being impactful or steering debate)” (1). This tension was also illustrated by Gade and Lowrey, who claimed that “journalism now is more focused on technology and associated skills, shifting the emphasis of journalism work from creative inputs (news gathering, creative enterprise, verification, writing, editing, and design) to outputs of content in multiple platforms and products)” (2011, 33). As Eldridge (2018) observed, reliance on digital technologies in the newsroom has forced journalists to reconsider how they define themselves and those they work with while allowing them fewer opportunities to cast themselves as journalistic “insiders” and everyone else as “outsiders.”

Web analytics companies have their own occupational and ideological foundations that differ from journalistic norms and values. Because web analytics companies provide a for-profit service to news organizations (i.e. they sell web analytics to journalists), they have worked to understand journalists’ norms and values while appropriating some of the tenets of journalism to fit their own. Scholars have argued that such mingling between news professionals and non-professionals can be coordinated through an understanding and possible exchange of norms and values (cf. Eldridge 2018; Lewis and Usher 2016). Petre (2015) wrote that using “dry and technical language,” web analytics companies present themselves as able to help clients [including news organizations], to equip them with “decision-driving data,” and to show “what content sparks and holds readers’ attention.” “The prevailing message” added Petre, “is that the dashboard exists to communicate rational, dispassionate data, upon which journalists can then act” (16). In seeking to coordinate with journalists, Petre suggested that “designed-to-communicate data” emulates “traditional journalistic values and judgment; it must be compelling; it must soften the blow of bad news; and finally, it must facilitate optimism and the celebration of good news” (17). She added that Chartbeat staff would defer to journalistic judgment in verbal and marketing communication, build trust with metrics, not make recommendation about what journalism should be, provide “tough love,” and act as facilitators and rationalizers of numbers in newsrooms (17). Petre highlighted a conscious coordination between journalists and web analytics companies centered around news production.

The concept of boundary maintenance around boundary objects suggests that professionals and non-professionals can work together to communicate and collaborate toward a shared goal (e.g. news production) (cf. Lewis and Usher 2016; Star and Griesemer 1989). A long line of research on digital journalism has confirmed that traditional news organizations have sought to accommodate themselves with the changing media environment from adaptation to technological convergence (Cottle and Ashton 1999), the web (Boczkowski 2010), web metrics (Anderson 2011a), and digital and social media (Belair-Gagnon 2015; Eldridge 2018). In examining the role of programmers and journalists in news production, Lewis and Usher (2016) found that “different parties may readily recognize the general contour and importance of news [the boundary object] while having different mental models for what the internal workings of news (should) look like” (548). The same has been found of journalists intermingling with web analytics: journalists value analytics in helping to determine and drive the production of news and engagement with audiences, but they do so while also considering their journalistic intuition and traditional norms and values (Petre 2015). How exactly web analytics see themselves in the same process remains relatively unknown.

Journalism as a profession is the complex and intertwined result of reactions from in-group and out-group dynamics, and how they differ from “internal faults and failing” in boundary maintenance (Eldridge 2014, 2). Journalists are autonomous and controlling of what stories to cover and how to cover them. They see outside contributors as interlopers, or unwelcome guests who maybe providing a service to the news process but are not themselves journalists. Scholars have already begun exploring how journalists perceive and engage with outsiders, or interlopers, but literature on how the creators of that data (i.e. web analytics companies) position themselves within journalism is limited (cf. Petre 2015).

Eldridge’s (2014, 2018) concept of interloper media is useful to analyze web analytics companies’ engagement with the broad boundary object of news production. Boundary objects are those objects, concrete or abstract, that carry different identities and meetings in different social or professional arenas and can help those different realms communicate and understand one another (Dodgson, Gann, and Salter 2007; Star and Griesemer 1989). In journalism, boundary objects can be represented in a variety of forms ranging from specific news topics that connect journalism with multiple audiences (e.g. science journalism engaging general news audiences, policy-makers, scientists, and those interested in science) (Shanahan 2011), evolving forms of journalism such as “localism” that connect audiences in new and significant ways (Dunbar-Hester 2013), and the process of making news wherein editors and reporters work collaboratively through a unifying exercise to create and disseminate the news (Anderson and De Maeyer 2015). In the case of news production, web analytics companies have stepped in to provide a service that aids in the process of news selection, helping them collaborate with news organizations toward a common goal in a way that suggests they are less media interloper and more welcome guests.

Despite anxieties expressed by journalists toward web analytics, web analytics companies provide services that have the potential to enhance the news process. These companies have limited their encroachment on journalism by not claiming to be journalists, rather casting themselves as an aid and non-antagonistic to journalism. News organizations are often not the primary clients for web analytics companies, advertisers, or other companies are. As Petre (2015) suggested, web analytics companies have not historically claimed journalistic values and practices. They have been deferential to journalistic norms and practices while selling their services. In other words, these companies do not purport to be part of the in-group; they are in the out-group that happens to have a connection to journalism through the boundary object of news production as well as an interest in the success of journalism.

Building on this literature, this paper seeks to go a step further in conceptualizing the roles these companies express having in news production in order to determine how they fit into changing journalistic boundaries and ideas of interloper media. Exploring these roles is central to understanding how external companies (not defined primarily as journalistic organizations) are influential in contemporary newswork and how they shape (and are shaped by) journalistic norms and values as well as by introducing and negotiating web analytics values in news production.

Understanding these companies’ perspectives matters since their pronouncements affects the news production process and can potentially change the public good ideal of journalism to a more tech industry oriented one (cf. Gade and Lowrey 2011). Rather than looking at practices, this paper focuses on hearing how managers at web analytics companies formulate and sell ideas about how journalism ought to work. While these companies affect the autonomy of journalism, they have not appeared to be doing so malevolently or in an effort to steal journalism’s professional jurisdiction. Rather, as previous studies have shown, the web analytics they supply and the means in which they are incorporated into the news process appears to be shifting some traditional journalistic practices. Thus, exploring how web analytics companies position themselves in journalism and how they maybe seeking to affect jurisdictional claims journalists make is important for understanding how the process of news production is changing and what role key influencers such as web analytics companies maybe playing (cf. Carlson and Lewis 2015; Lewis and Usher 2013, 2016).

Method

To approach the question of how web analytics companies understand and position their work in relation to journalism, we first identified influential web analytics companies working with news organizations. Drawing on the Reuters survey of global newsrooms, which indicated a small contingent of web analytics companies dominate the market (Levy 2016), we began by developing a list of the top web analytics companies working with newsrooms. We searched the names and emails of managers of the top 10 web analytics companies included in the Reuters survey. We then contacted these managers in March 2017 to take part in interviews, receiving nine responses from eight companies. We informed managers that we would ask questions about their approaches to and relationships with news content and news organizations. We conducted interviews with all managers from March to May 2017 including via phone or video conference. On average, these interviews lasted 45 min. We anonymized all interviews because some of these individuals wanted to remain confidential. As such, responses are listed in the findings as Manager A, Manager B, etc.

The interviews we conducted are elite because they represent the economic notables of a small group of businesses who currently dominate the market of web analytics in news production. As of late 2017, the web analytics companies we interviewed represented an influential contingent of web analytics companies working with news organizations. These companies were founded between 2008 and 2012, collectively employed more than 900 workers, had more than 50,000 clients, scanned more than 400 billion pages of content every month, worked with news media outlets across at least four continents, and received financial support from media-focused organizations such as the Knight Foundation, the McCormick Foundation, the Conant Family Foundation, New Media Ventures. This research has thus high internal validity since we talked to leading players in web analytics in journalism, a market that is driven by a handful of powerful actors. Though social media platforms such as Facebook and Twitter offer web analytics, they do not represent themselves as nor do they function primarily as web analytics companies. Similar to Carlson’s (2017) work, the goal of this research is to understand the how these companies communicate their role to journalists, as opposed to conflicts of these roles in daily practices.

Given that research is only beginning to develop in this area, we employed semi-structured interviews in order to help reveal how these managers understood and positioned their work in relation to journalism. Semi-structured interviews draw from a pool of questions and allow respondents to direct the flow of the interview in a reflexive format, often drawing out direct, indirect, and unexpected responses that may better represent respondents’ thinking than more structured questioning (Anderson 2012). Interview questions included: What is your understanding of the role of journalism? Has your understanding of the role of journalism changed over the past years? And if so, how? How is your work different from and similar to the work of journalists or other web analytics companies? How do you work with news organizations to implement change (what are the challenges and how they try to solve it)? How are products developed for journalism? Who decides what analytics are valuable in journalism? What training is in place for reporters? We also asked questions regarding trends looking forward the industry. We also asked managers to consider the role of their company in the newsroom and to reflect on other companies’ best practices.

Once we collected and transcribed the interviews, we combined authors’ notes to help yield unique and overlapping responses. We then developed these notes into role categorizations (Konieczna 2014). Following the suggestion of previous scholars (Birks and Mills 2011; Brennen 2013), this approach allowed for an open exploration of the perception managers had of web analytics roles as opposed to relying on preconceived categories of roles that, while potentially useful, could have restricted the findings.

Findings

At a time when web analytics are themselves becoming a boundary object in newsrooms, allowing web analytics companies, news organizations, and journalists to collaboratively harness the power of numbers for more efficient and engaging news delivery, web analytics companies appear to be casting themselves transparently as attempting to balance financial and public service goals. Managers interviewed for this study discursively positioned themselves both as experts providing a complex service and as working collaboratively with news organizations to achieve common goals.

This section discusses how these coordination efforts, built around the boundary object of news production and broken into roles, were articulated by the managers interviewed. While previous research has indicated a tension between journalists and web analytics (though not necessarily between journalists and actual web analytics companies), managers suggested that by coordinating their efforts with news organizations to prioritize existing journalistic norms and values, and by harnessing the boundary object of news production, they maybe easing existing anxieties while also making web analytics a collaborative and necessary tool for journalists. This coordination effort, outlined in more detail below, is reflected in the relationship managers said they seek to build with news organizations.

Managers said that when they first began working with newsrooms, they typically provided analytics to a limited number of staff. They worked primarily with editors who expressed a need to control the flow of web analytics through the newsroom. More recently, they indicated they are working with a broader array of journalists, including reporters, and attempting to be transparent in how their analytics are developed and shared. As Manager A observed:

One of our main values at the company, we’re very intent on transparency in measurement. We want the brand to know that the data they give us, that we’re transparent with how we measure it. The brands are starting to get nervous that metrics that are being shared with them aren’t truthful.

Managers also noted an uptick in interest and use of web analytics among reporters, who they said were more engaged with predictive analytics and seemed to be working to understand how metrics and analytics were gathered, processed, and implemented. Where working with web analytics used to mean “somebody emails me a spreadsheet every Monday and I pat myself on the back,” journalists told managers they were receiving detailed reports that they understood (Manager B). And they considered themselves as serving as proponents for journalists and publishers while enhancing the public’s connection to and engagement with journalism. As a manager puts it, news organizations are no longer using pageviews, shares, or likes as a primary mean to gage the success of story, but are instead using a variety of analytics “to bake [them] into their editorial strategies right at the content creation stage” (Manager A).

Journalists may still use analytics as indicators of content reach or popularity with audiences on news sites and their respective social media platforms, but they use web analytics that are more geared toward predicting outcomes early in news production or even before that process begins. According to managers we interviewed, such acceptance of analytics represents a paradigmatic shift in the relationship between journalism and web analytics companies from one where the latter were viewed as encroaching outsiders to one of coordinated efforts built through the connective power of news production.

As the subsections below illustrate, the managers interviewed here saw news production as a boundary object (though they did not use that phrase) wherein they could connect with news organizations to better understand and make use of the journalistic norms and values that are core to the profession. These efforts could be categorized into three main coordination roles that managers said they sought to establish with journalists: disruptors, connectors, and routinizers. These roles are professional constructs, are non-exclusive, and overlap with each other to provide insights into how web analytics companies are using news production as a boundary object to enhance the news process while also making themselves integral to the future of journalism.

Coordinating News Production: Disruptors, Connectors, and Routinizers

Disruptors

In positioning themselves as disruptors, managers placed themselves as providing new insights into news production, though they remain met with some skepticism. As Manager C said, news organizations seem like they wonder if “we are selling snake oil. Look at these numbers because they are numbers and they will help you!” To help shed journalists’ anxiety, managers said web analytics companies’ primary goal is to solidify themselves as a necessity to journalism by providing disruptive analytics that help themselves and newsrooms understand news audiences while creating a more sustainable business model for news organizations. As Manager D said,

We think [we are] kind of disrupting the data science for publishers and removing the need to spend a lot of time, because obviously, there is all the other things in the newsroom like reporting the news. That’s what they want to focus on, and that’s what the we do to add the most value … we help them find new ways to distribute their content.

While managers confirmed web analytics were still emergent innovations in journalism and already causing new organizations to rethink their approaches to news coverage (i.e. “journalistic instinct is no longer enough to warrant resources for a story”; Manager B), they noted that web analytics are disrupting the goals news organizations traditionally set, how they achieve those goals, and with what resources. As Manager A highlighted, web analytics companies worked early on to understand how journalism worked and what goals each individual news organization wanted to achieve. But as their understanding of journalism evolved, and as they improved their capability to predict when and where audiences might be, they helped reshape the motivations behind news coverage and placement.

While managers said they were helping news organizations create more achievable and sustainable goals in terms of audience reach and economic viability, they were almost exclusively more interested in their role as drivers of content based on audience web engagement and journalists’ ability to control the fate of that content. Managers also said they were part of shifting journalistic practices influenced by web analytics culture that increasingly includes experimenting with new technology (e.g. apps), new forms of constructing narratives (e.g. to fit with mobile consumption), and new ways of approaching data. As Manager B indicated,

You have to marry the qualitative and the quantitative. I don’t think it’s a bad idea to have experimentation and collaboration, whether it’s an editor or somebody through social media or whoever is manning the project or measuring the success of the project.

According to managers, while journalists were initially hesitant to use web analytics, or were restricted from accessing them by their news organizations, they are more often seeking ways to access, understand, and apply such data. “Is there a way,” Manager B posited, “that you can recycle this content or learn something new about your audience? That’s sort of constant feedback, that constant reevaluation is what I hope. That’s the trend that I hope we are moving towards.”

Managers indicated they seek to enter the news market in order to support their own companies’ financial goals. Rather than forcing newsrooms to modify already volatile budgets to make room for a host of analytics-focused employees, managers said they have emphasized the simplification of data collection, processing, and contextualization so that resources are not as burdensome as they previously were. According to managers, this is increasing the value of web analytics, while also challenging news organizations who have not yet embraced analytics to incorporate them into their editorial decision-making at the very least. Such pressure, or “opportunity” as described by Manager E, points toward a future where news organizations have access to web analytics and the ability to combine those with existing journalistic instincts when considering what content to cover for which audiences. This disrupts the traditional news process in which journalists have experienced limited access to web analytics and have leaned on their journalistic instincts for content, providing a more tangible pathway (i.e. web analytics) for content selection and placement while indicating that web analytics companies are invested in the boundary object of news production.

Connectors

The connective, or collaborative, ability of web analytics companies surfaced repeatedly among managers, who reported that web analytics companies seek to work with news organizations and journalists to improve connectivity with audiences while also connecting analytics companies’ expertise with other professionals in the newsroom who might influence news production. As explained by Manager G:

A couple of years ago [redacted] had a huge story where, using [redacted], they saw a story about a schizophrenic teenager who had been insulted by police or something like that, on a local news site in a local blog that didn’t have a big audience that had gained traction. [Using our web analytics], the team of editors and journalists were able to see that that was something that would resonate with their readers. Then they were able to do their own reporting and add some additional context around it and bring it to their much wider national audience. That was a phenomenally successful story for them.

According to managers, in helping connect editors and journalists with analytics that indicate or predict successful content, often in real-time, web analytics companies are helping news organizations better understand the behaviors and potential needs of news audiences. For example, a manager explained how his company used a video product to help a publisher understand opportunities to “push out new videos and stories that might might benefit from video” in ways that increased the amount of time audiences spent with content, and allowed journalists to view in real-time a more effective way of engaging with viewers. Following this exercise, according to Manager E, the publisher:

[H]ad some enormous increase in their video audience over their first few months of working with the company, so it really [had] a real revenue ramifications. And there are a few great examples like that of sort of businesses finding new revenue and new life in their audience by using [the company’s] tools to understand which content they produce resonates with which people.

In this way, web analytics companies put news organizations in experimental analytics environments by linking daily journalistic practices with web analytics companies’ financial goals. That is to say, managers indicated that as they worked to improve the collection and delivery of web analytics, they frequently tried out new approaches and tools. They worked simultaneously to connect journalism to sustainability in commercially driven newsrooms, specifically noting a change from the top-down in terms of how web analytics are improving digital connections with audiences, while also trying to improve their own products.

Routinizers

Such tinkering is part of what managers said was a new form of routinization in newsrooms: one wherein tools such as web analytics were expected to change rapidly to meet the evolving demands of news organizations. In positioning themselves as routinizers of web analytics into journalism practice, managers indicated that web analytics companies work within the context of news organization structures and norms and values on a daily basis, relying on a constant stream of web-based data, or feedback as explained by Manager E:

[We] help people sort of ease the transition to digital. One editor referred this to me in kind of the most eloquent way is forcing editors to confront the fact that they have an audience. So when you are looking at [our] dashboard, because the numbers are alive, it’s sort of suddenly in this very human way makes you realize, “like wait, there are other people there right now reading my stuff, and depending on what I write or what promotional decisions I make, like different people will read different things.” It really kind of humanizes traffic online in a way that has really helped people to shift to think about their audience on digital.

Yet, even when speaking in terms of working within the news process, managers still positioned themselves somewhat as outsiders working for and within newsrooms without making final news production decisions. By providing web analytics in real-time, and by ensuring those analytics were received and processed by editors, managers, and reporters, managers said their companies were coordinating with journalists to make web analytics more routine.

This kind of coordination effort in routinizing web analytics within news production took multiple forms, including: training, support calls, webinars, corporate statements explaining particular processes and metrics, and routine check-ins with news organizations to see if all needs were being met. Managers mentioned they also have worked to routinize web analytics by simplifying data and the tools used to provide it (e.g. dashboards that are easy to understand), being readily available to help with projects and dissect data alongside journalists, or meeting in person with news organizations to explain how data could be used to achieve certain goals.

Managers indicated that by designing algorithms and toolkits tailored to individual news organizations’ needs, they have been able to make themselves a reflexive and routinized part of the news process that news organizations used for everything from forecasting which content will work best with what audiences to on-the-fly adjustments to headlines, photos, audio clips, colors, filters, and even single words embedded in stories to enhance their penetration with audience across news platforms (e.g. newspaper, desktop article, mobile article, social media snippet). According to Manager B, journalists are learning lessons from the incremental adoption of social media and been much quicker to welcome analytics into their journalistic repertoire.

Web analytics companies thus may develop processes with news organizations to implement web analytics and to achieve a set of goals that both integrates web analytics expertise with journalistic norms and values. Because much of the data provided to news organizations by these companies is live, meaning audience interactions can be seen as they happen, newsrooms are able to work with companies in real-time to make sense of the data and quickly make adjustments. As Manager E observed, such 24/7 accessibility forces editors

to confront the fact they have an audience. So when you are looking at a [redacted] dashboard, because the numbers are alive, it suddenly in this very human way makes you realize “wait, there are other people there right now reading my stuff, and depending on what I write or what promotional decisions I make … it humanizes online traffic.

Managers said that journalism is an institution built on core, routinized values such as objectivity, truthfulness, public service, promptness but that news organizations are wrestling with how to connect with news audiences in meaningful in sustainable ways. Web analytics, they said, are a sustainable and necessary means to understand what content, or pieces within content (i.e. headlines, photos, links), is or will be successful. By routinizing them into news production, news organizations can make quick changes and decisions with immediate results. As explained by Manager E, this may include the tweaking of a headline to reach a particular demographic or the development of an interactive video to go alongside text because web analytics predicted a broader reach with that approach.

These changes and the reasons behind them, managers suggested, should be transparent for journalists, and perhaps audiences, so that they might more quickly become accepted and trusted as part of an evolving news production wherein web analytics can be used to improve content and content placement. By encouraging news organizations to be more transparent about web analytics with journalists, and journalists to be more transparent in the application of those analytics with audiences, managers revealed they considered themselves part of a routinization of analytics in news production. As Manager E remarked, “these days, [web analytics] tend to be a lot more open and shared within the organization. That sort of internal transparency conversation has changed and moved toward one where there’s a kind of open and free access to analytics.”

According to managers, that freedom affords journalists the opportunity to experiment with analytics without the worry of constant oversight or the divulging of what once was considered sensitive and restricted data, thus making them more routinized into news production. As described by Manager E, “it is taking the numbers and actually finding opportunities that sort of moment by moment do things to improve the numbers.” In other words, as cast by managers, web analytics companies provide an opportunity for web analytics to enhance metrics of success while also creating an atmosphere wherein they can be accepted as a useful and routine part of the news process.

Summation of Findings

Each of the above roles positions web analytics companies as efficiency drivers, successfully using the boundary object of news production to bridge their goals with those of news organizations. As Manager D stated, the key is contextualizing what efficiency is for each news organization without purporting to actually be that news organization or a journalist of any kind. To that end, managers positioned themselves as close to, though not fully inside of, the journalistic business model that news organizations and web analytics companies benefit from.

Managers placed their companies’ roles in news production as increasing the efficiency of the news process and journalistic outcomes, taking pleasure in driving and create revenue streams, as well as in improving the impact of content almost instantaneously. As Manager F described it,

In some cases, you might see this really important story that’s not getting a lot of traffic, and that identification that it’s not getting a lot of traffic or engagement signals that we need to do something different with it to promote it in the places that it can get the engagement that we show it deserves. So focusing on that distribution and creating even more value for the work that people are already putting in is really important for us.

Along with real-time changes to content, managers said that by offering a 24/7 flow of analytics coupled with analyses and interpretations of that data, web analytics companies are enhancing the effectiveness of news content. As put by Manager B,

The ability to realize to the second “Oh there’s a problem,” and immediately take action and immediately see whether the action you took is reflected in the numbers, let them make changes, and immediately fix the problem on a really important story that otherwise may have not reached the audience that it did.

These managers’ understanding of their position relative to journalism should not be dissociated from their interdependence to the boundary object of news production. At the same time that these managers expressed how they control and deploy their resources, they saw themselves working within a process driven by a fluidly changing journalism profession. Reflecting the work of Petre (2015), these managers were positioning their roles in news production as both oriented toward themselves (business) and toward the journalistic profession (public service), indicating a broad coordination effort on their part. As Manager F described it, “we are developing an approach built on efficiency that is repeated and returned to over and over again” in a way that creates “loyalty and bridges news with numbers.”

Conclusion

From the perspective of news organizations, the coordination efforts used by web analytics companies are deeply felt as journalists rely more on web analytics while assuming the risk of commodifying audiences at a time when they wish to be humanized while loosening the professional autonomy that journalists prize (cf. Duffy, Ling, and Tandoc 2017). By allowing audiences (and their predicted behaviors, aka web analytics) to guide editorial decisions, journalists may risk exchanging what is and should be news that serves the public interest for only that which seems interesting to the public. As our findings indicate, web analytics companies have taken note of such concerns, working to position themselves not as unwelcomed collaborators or media interlopers but rather as analytic advocates who recognize their own companies’ success correlates, at least in part, with journalism’s success. To that end, web analytics companies are working to ease the complexities and layers of today’s journalism without actually practicing journalism. They see and promote, the benefits of the web analytics they provide but do not purport to be drivers of cultural or institutional change among newsrooms or journalists.

As web analytics companies occupy a greater role in the news process, particularly through the boundary object of news production, they may see their position as an emergent and integral part of journalism. Our interviews suggest that web analytics companies are external to journalism and wrestling with self-perception while also relying on the production of news to learn about and connect with the practice of journalism. Web analytics companies may not necessarily claim responsibility for any of the changes their involvement may have on journalism culture, but they nonetheless are having an effect on workflow and how journalists innovate in news production through the use of web analytics. They are, as Eldridge (2018) notes, agents of change in journalism, questioning their own identity while reshaping the identity of journalists and journalism.

The findings of this study suggest that web analytics companies coordinate the boundary object of news production, as well as roles between web analytics and journalism, by organizing and working with journalists on real-time data about news audiences and news content. The interviewees related these practices more to economic viability rather than news production values (e.g. journalistic objectivity and news judgment). In other words, managers shared the idea that web analytics is for good journalism to the extent that it can enhance the visibility of good content and can form a sustainable business model for news. And when business is good for news organizations, business is good for web analytics companies. They can use such positive trends as indicators of their role in news organizations’ financial successes. Their acts of coordination highlight their boundary work (experimentation, web analytics, and profit oriented) with boundary objects (news production values and norms) in ways that, while keeping the tenets of journalism in mind, still serve the primary objectives of their companies.

The larger finding of this article is that while web analytics companies seek to understand and address news production values and norms without assuming responsibility as journalists, they foster profit-oriented norms and values in newsrooms by introducing web analytics as disruptive, connective, and routinized in news production. By offering a product that they (i.e. web analytics companies) need to modify on a continuous basis because of changes in the structure of the web and audience behaviors, they foster a milieu of constant experimentation with old and new products. These findings reflect some of the academic work produced on the effect of technology industry and social media platforms on journalism, including programmers–journalists collaborations (Lewis and Usher 2016), news distribution (Bell and Owen 2017), as well as journalistic sourcing (Belair-Gagnon et al. 2018). These companies recognize and navigate what so many scholars have observed: journalists temper the potential value of web analytics with an anxiousness driven by inexperience with data and concerns over the loss of autonomy (cf. Petre 2015).

Our approach offers a thematic foundation for future studies, which could use additional interviews with managers and direct observation on how these companies work with news organizations. Also, as these companies develop (in terms of technology, precision of analytics, connections to social media companies, etc.), there is an opportunity for sustained quantitative and qualitative research that explores the role of web analytics in journalism and how they differ across contexts (e.g. crisis reporting, beat reporting, non-profit v. for-profit news organizations). Such research would provide scholars and practitioners with a deeper understanding of the effects and nature of interactions between these companies and journalism.

Disclosure Statement

No potential conflict of interest was reported by the authors.