For Grupo Reforma, understanding why users churn is a top priority, said David Hinojosa, IT research and development subdirector. The company, which publishes 10 daily newspapers in five cities in Mexico, is counting on this knowledge to help increase subscriptions 15% by 2022 and 60% by 2025.

“With that information, we think we can make more informed decisions to grow our base,” he said.

The hypothesis of the project states that developing both qualitative and quantitative tools to identify reasons for churn will help the company transform content and/or the way it is presented and discover how relevant its subscription offers are.

Quantitative tool

For its quantitative tool, the company aimed to:

• Build a data model based on the navigation and conversion from those subscribers who churn.
• Discover the behaviour patterns they had before their churn as a warning sign.

The company defined six variables for its quantitative tool: frequency, content consumed, categories, concepts, freshness, and duration.

“With those six different variables, we tried to find a relationship between those variables and engagement levels,” Hinojosa said.

Using the first version of its engagement index formula with these variables, the team can chart a correlation and identify the propensity to churn. The team plans to continue A/B testing and to improve its data model moving forward.

Qualitative tool

For its qualitative tool, the company aimed to:

• Find out directly from former subscribers why they churned.
• Organise, categorise, and count the factors/reasons.
• Define strategies and actions to reduce the churn.

Grupo Reforma sent surveys to former subscribers to discover reasons behind their churn and offered 14-day free access for those who completed the survey. More than 70% of survey takers used the trial.

The survey itself used open-ended questions so users could describe why they churn. It also asked for suggestions to improve products and user experience. The company aims to use this tool to get insights daily so it can continue to improve its offerings.

Although both the quantitative and qualitative tools are only in their first version, Hinojosa said they helped Grupo Reforma prove its hypothesis: “We did it. We found most of the important reasons why users churn and we are taking actions to improve that churn.”

This case study originally appeared in the INMA report, The Benefits and Risks of Media Data Democratisation.

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