Irish Times analytics team plans for paid content model, deciphers reader behaviour

To achieve growth in audience and revenue, The Irish Times needed a better understanding of its readers. This identified the need for a new function to process and analyse data effectively, as well as to deliver actionable insights to enable informed, data-drive decision-making within the news media company. 

A new business analytics function was set up to provide decision-makers (including editorial, marketing, sales, circulation and finance) with the best available intelligence about readers, customers, revenues, operations, suppliers, and the market in a timely manner. 

With the planned introduction of a paid content metered model on within one year of the set-up of the business analytics team, The Irish Times prioritised key insights to support this initiative. The achievement of this goal would be identifiable by the following behaviours within the organisation: 

  • An increased reliance on data and analytics by decision-makers who consider it critical to their competitiveness. 

  • An increased level of output of business intelligence and analytical solutions that have a high level of influence on employees’ actions.

The Times first visited and spoke with a number of market leaders in this media space to identify best practices in the use of analytics. It also recruited the services of a specialist consultant in this area who worked with us on site for a period of time to help develop this new function. 

A new central business analytics team was set up in The Irish Times, incorporating existing research specialists, as well as newly recruited data analysts and a search and analytics specialist. A dedicated insights meeting space was created to encourage communication of actionable insights and ideas within the business, and a series of staff briefings were held to introduce the new function. 

In addition, a new daily content performance report is circulated to key staff each morning by 10 a.m. by the analytics team, giving easier and broader access within the company. A member of the analytics team attends the daily editorial planning meeting each morning to provide updates and insights on reader behaviour, as well as to stay informed on upcoming developments. 

A series of Web-based content dashboards have been developed in-house. They are updated daily and report on content performance by site section, by author, by day, by referral source, and other useful data points. 

A Web-based subscriber analyser dashboard was also developed in-house in time for the February, 23, 2015, launch of a paid content metered model on A Web-based subscriber analyser dashboard was also developed in-house, updated daily. This reports on the number of new subscribers, churn levels, content preferences of subscribers, etc. 


  • Through an improved understanding of reader behaviour and content preferences on various channels, more informed decisions are made around the best time, day, and channel to push particular Irish Times articles, leading to better optimisation of content to extend its reach. 

  • Web site traffic grew by 22% year-on-year, from 2013 to 2014. 

  • Social referral traffic in particular grew by 45%, continuing to grow further in 2015.

  • Through communications and increased awareness of the business analytics function within the company, the team is involved on a regular basis in internal meetings to help inform decision making, planning, and experimentation. 

  • There is a marked shift toward a more frequent use of data, with easier access to insights available to a broader audience within the company.

  • A paid content metered model was implemented on February, 23, 2015, with key elements informed through reader analysis. With the business analytics function in place, The Times are in a strong position to quickly analyse, react to, and deliver insights on changes in reader behaviour as the model progresses. 

The Times identified four key components to be put in place: 

  1. Data: What data was available? How could we put it to use?

  2. Technology: What systems needed to put into place to process and analyse this data? 

  3. Team: What skill sets were needed? 

  4. Insights: What were the first key insights that were needed to garner to benefit the business? When a planned introduction of a paid metered model within 12 months, insights to support this were a primary focus. 


The goal was to build from user behaviour on and prioritise key data points to support a paid content initiative. Before the establishment of the business analytics team, the main source of data on Web site behaviour was Google Analytics, which gives aggregate data of reader behaviour at the group level.

The biggest data sources available that were not actively in use were The Times’ Web blogs, which record every interaction on This was the obvious first data source to work on bringing into a format that could be analysed. 

By also incorporating its content management system, this then allowed The Times to link an article’s performance with the time it was published, the day it was published, the topic, the author, etc. The aim here was to bring these available data sources together. By joining them up, The Times could gather more insights from them than could when they were separate sources.


We needed the right size technology to suit our needs. We have used open-source tools where appropriate to manage costs and to allow us to be flexible as we worked through our needs, but at all times making sure that the technology can scale to meet future requirements.

Cleansing the data has been a vital part of this process, ensuring the accuracy of our outputs. 

The data is then pushed through to our data warehouse, where we query the data and release our findings in various formats, including Excel reports and internally developed, Web-based dashboards. 

Our technology design principles included: 

  • Getting the right size technology to suit our priority needs. 

  • Using open-source tools where appropriate.

  • Ensuring our technology solution can scale to meet future needs. 


We wanted to create a central team to serve the business, assign dedicated office spare to support business insight sharing, identify key roles, and recruit to support business requirements. 


  • 85% are occasional readers. 

  • 8% are frequent readers. 

  • 7% are heavy readers. 

Other than pageviews, how else could we measure reader engagement?

One of the first new metrics we reached when we started building this function was a measure on engaged readers. From analysing our Web blogs, we established that there was a consistent pattern each week on the percentage of readers falling into particular segments, which we’ve categorised as occasional, frequent, and heavy readers, based on the number of articles they read. 

  • Occasional readers: Unlikely to pay for access to content, but they are, of course, an important segment of our audience and vital for advertising purposes to generate traffic. 

  • Heavy readers: The most loyal segment, and the most likely to pay for content once the metered model on launched. 

  • Frequent readers: A group of readers we could attempt to convert into “heavy” readers. 

This analysis informed our decision on where to set the meter on, with the end decision being after 10 articles per week on our Web site and after 20 articles per week on our mobile site. 

After further analysing these reader groups, we noticed some common trends. For example: 

  • Our occasional readers were more likely to come to the Web site via external sources, such as Facebook, Twitter, and Google.

  • Our loyal readers were more likely to visit directly via the home page, and to consume more articles per visit. 

  • A higher percentage of our heavily engaged readers were from Ireland, compared to a greater mix of international sources in our occasional reader category. 

  • A high percentage of our heavy readers were engaged with core Irish Times content: sport, business, and opinion. They were also more likely to read our regular columnists than our occasional readers, who had a higher propensity to read breaking news stories. 

These insights helped inform our decisions around content strategies as we planned for the launch of our paid content metered model. 


We developed a series of Web-based dashboards for editorial, designed to report on our content performance and help us optimise our content by showing us the best time and channel to publish/promote particular content.

About L. Carol Christopher

By continuing to browse or by clicking “ACCEPT,” you agree to the storing of cookies on your device to enhance your site experience. To learn more about how we use cookies, please see our privacy policy.