A Gazeta’s Web site is just one distribution point for content for the Brazilian news media company. Others include videos, podcasts, newsletters, WhatsApp, and social media platforms such as TikTok, Facebook, and Instagram, which make up the group’s ecosystem.

“The measure of success is not just about driving traffic to the Web site,” explained Aldo Agurto, marketing intelligence specialist at Rede Gazeta. “We have to understand that the audience should have a complete experience with any of the products or with just one in particular.”

Rede Gazeta is the largest media group in the state of Espíritu Santo, Brazil. It started operations 95 years ago and includes 14 brands, including TV, radio stations, news portals, and the A Gazeta Web site, which covers current news and entertainment.

With four million users, eight million pageviews, and three times the traffic figures of its nearest competitor in the state, according to Comscore, A Gazeta has turned into a success story.

In a recent INMA Webinar, A Gazeta, a journey to use data in digital journalism, Agurto explained that the way the company handles data is different from competitors, he said.

Beginning the digital transformation

A Gazeta’s current digital project was born in 2019. The media group already had a printed product and a Web project with a different name. It was a gamble to focus solely on digital, a move that was always supported by the media’s management.

The initiative accelerated with the arrival of COVID-19 and involved launching a completely new project while ending the production of the print newspaper.

This was a new content offer and new distribution channels for audiences. “A new moment, a new episode, a new focus on the transition to digital,” Agurto said.

The initial digital focus was on the audience in the state of Espíritu Santo, based on two pillars: Offering high-volume, evergreen, soft content alongside a more exclusive content proposition for subscribers. In other words, two products under the same brand. They also identified two different consumption moments, two types of behaviour.

After that process, a decision about what technology should be used needed to be made. Was the smart move to expand their existing technology? They decided to continue with Piano technology and see what it offered when combined with data management.

There was no data history beyond a year due to contract limitations, so they decided to make adjustments to expand data history.

For this reason, they implemented Piano Insight, Piano DMP, and Piano Composer: “We designed an architecture that allowed us to harvest data, work with more information, and link more contacts,” Agurto said.

This allowed the input and capture of information by Google Analytics. The combination of Piano and Google Analytics let Piano focus on journalism while Analytics focused on the digital advertising market.

These were accompanied by a data visualidation architecture based on tools such as Data Studio, Power BI, Chartbeat, and others like Hotjar, which allowed them to detect audience behaviour within the Web site. They also used CrowdTangle, which monitors everything happening on social media.

This architecture changes along with business needs, while the data architecture adapts to analysis needs. “Everything is in constant development,” Agurto said.

The current data architecture includes subscriber databases, other products such as TV, Web portals, and radio stations, as well as Google advertising.

They also created an internal data analysis portal so employees could access this information from anywhere based on their profile.

This entire process was outlined in a road map that included starting with simple data discovery, integration, and analysis, and then moving onto advanced analysis. This road map is constantly evolving and progressing.

One key aspect is that the analysis and data management had to involve journalists, editors, administrative staff, and commercial and IT teams, Agurto said.

Data with a purpose

Agurto pointed out that it is necessary for the collected data to be transformed into “information, knowledge, and intelligence” that can be easily understood, emphasising this is a slow but continuous process.

In these processes, it was also necessary to adjust the human team and analyse the profiles of the analytics team, while also relocating those individuals to different tasks.

All the generated information had to flow in the best possible way for it to be understood by the different teams within the company.

Based on this analysis, they reviewed some internal movements, benchmarked with local and international newspapers, gathered data, identified needs within the company to define its relevance, and conducted internal audience workshops, worked with subscribers, and so on — all while considering that these processes should be done periodically to look at metrics, monitor new things, continue acquiring data, and conduct analysis.

Based on internal movements and information recycling, they proceeded to understand the audience, although they continued with the usual practices of measuring the number of users, how many came from computers and how many from mobile devices, how many were organic, and how many came from social media, although with an innovative approach.

What changed was the way they analysed these audiences. They began to delve deeper into this knowledge through initiatives such as heatmaps, segmentation, broader historical data. They focused on relevant information and what worked and what didn’t.

They also conducted studies and analyses, some of which were relevant while others were set aside. “The key to all of this was leaving fear behind and taking the steps that allowed them to start,” Agurto said.

Audience segmentation

From then on, with a lot of information available, they proceeded to segment audiences by “respecting their differences and similarities to gain a deeper understanding of their content consumption behaviour.”

They recognised users who were not registered, those who were, subscribers, those who surf shallowly, and those who explore the portal more intensively, among other variables.

From this attempt to segment audiences, questions arose: What is the moment of consumption for our audiences? Are there noticeable differences between each group? Is behaviour cyclical month to month? Does segmenting in this way influence our strategy? Does it help us grow our knowledge? Is this the best model or should we start with a simpler one?

As part of the goals set, they aimed to convert non-registered general audiences into subscribers.

This, in turn, was the beginning of setting up a paywall, which continuously changes as the media outlet gets to know the audience better and learns the tools to implement navigation barriers within the site.

To do this, the company appointed employees knowledgeable of numbers and the business to carry out the project.

At this point, they were able to segment users and conduct their respective analysis: light, medium, and heavy users, and subscribers. Each of these metrics should include KPIs.

Information for all

The next step was to democratise the growing amount of information among the different teams in the company.

They did this through routines consisting of weekly reports, monthly meetings, technical reports, and real-time systems. This information had to be shared on BI platforms. An internal Web site allowed authorised personnel remote access through dashboards and different profiles with information from digital, TV, and radio.

“It was a self-service of information,” Agurto said.

Additionally, they conducted studies on graphic interfaces, content evaluation with readers and studies to bring together different user profiles through surveys and heatmaps.

This feedback allowed the editorial team to make some adjustments and the technology team to monitor what was happening within the portal.

A control tower

From the previous exercise, the idea of creating a performance team arose: A multidisciplinary team that brought synergies and worked in the same place to analyse and gain knowledge about content consumption.

The group included various profiles, such as SEO experts, marketing professionals, editorial staff, and subscription specialists.

These exercises allowed the company to learn some lessons. For example, analysing and testing what works and what doesn’t and abandoning what is not relevant; understanding user routines within the portal; having an analytics team that understands data, technology, and statistical concepts; having a robust and flexible data infrastructure; integrating different players in the company (journalists, marketing, subscriptions, etc.); and creating routines to share results through meetings, reports, and presentations.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .