Colombian national newspaper El Tiempo launched its subscription model in 2020. Diego Barragan, head of advanced analytics, said the company is working toward increasing the conversion rate of its content behind a paywall.

“We wanted to increase our number of sales, and we started using an algorithm that would show what kind of content we can close and what can be more successful,” Barragan said.

Editors were closing certain content and giving away other content for free without having relevant data to back up their decisions, he said. Now, by using a machine learning algorithm, the company can estimate the success of certain news content and decide whether it belongs behind a paywall.

Barragan said the team created a user-friendly tool for editors that can filter by date or estimated success.

“We load the content, and then base decisions on this content,” he said. “This is how we make it available on the dashboard for the final user, for example, the editor or the journalist.”

The team at El Tiempo has validated its first hypothesis, finding that using a machine learning algorithm will help determine what articles will be more successful behind a paywall. It also knows the tool it built for journalists is useful to see what works best for subscription conversion.

“This allows them to see what are the most important articles in terms of the project,” Barragan said.

The next step is to run A/B testing to measure the impact of the conversion using suggestions from the model and conduct training to improve the accuracy of the algorithm. He said they hope to soon build an algorithm that will allow them to customise locked content for each user.

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

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