AI technology the lifeblood of new media

As Kenyan news organisations explore reader revenue as a potential avenue for media sustainability in the future, the adoption and optimisation of artificial intelligence (AI) becomes a critical conversation. Globally, the adoption rate of AI in the media is relatively low compared to other industries.

According to the 2021 World Press Trends report by the World Association of News Publishers (WAN-INFRA), 77 per cent of news publishers identified AI as a significant success factor in their business in the next three years. However, 70 per cent of media leaders said they had not yet fully optimised the benefits of AI.

We know that AI has been used for robot journalism, personalised article recommendations and automatic moderation of comments. Another growing area of research that could be relevant in the paywall era is how AI is being used to support reader revenue.

A recent report released last month by WAN-INFRA in partnership with German consulting firm Schickler notes that publishers are now very keen on using AI for reader revenue in four profound ways: prediction of conversion-likelihood, prediction of churn likelihood, market-based pricing and individualised paywalls.

There’s a fifth use case; the use of chatbots to enhance customer service, but that’s been around for years now, and editors found it the least relevant use case for AI in terms of reader revenue – at least for now.

A classic example of a news media organisation that has used AI to turn around its reader revenue business is the Globe and Mail of Toronto (Canada), which created an in-house AI solution named ‘Sophi’. The tool analyses consumers and content consumed to allow the newspaper to know whom to display the paywall to and whom to leave alone so as to encourage engagement with their content.

Litany of metrics 

In other words, Sophi will trawl their news sites and scan your audiences’ habits and is able to tell you, “Show this user your paywall because they are likely to pay” or “This user is not yet ready to make a commitment, give them time”.

In a world where news editors are bombarded with a litany of metrics from the number of pageviews, to the duration of the time spent by a user on your site to the number of visits, it is difficult to identify the metrics that matter and what exactly constitutes audience engagement.

One newspaper group in Norway, Amedia, is using AI to help them crack the code of audience engagement. They are using a machine-learning tool that crunches user consumption behaviour statistics on their behalf to come up with what the newspaper calls ‘Engagement Index’.

This engagement index is what Amedia uses to gauge if there are satisfying media consumption habits. It also alerts them to areas where their content is performing poorly, prompting them to rethink their content strategies.

The good news about AI use in media is that it is no longer the preserve of the big boys. Small media enterprises were found to be getting into strategic partnerships with other large publishers and institutions to create homegrown AI solutions.

The writer is the Director, Innovation Centre, at Aga Khan University; [email protected]