Looking for a new podcast to listen to? Acast wants you to skip browsing categorized lists and try the new Recommendations feature in its podcast app instead.
It uses machine learning to understand each user’s interests and listening habits, and then surfaces podcasts that they might like. According to the company, that goes above and beyond simply displaying a bunch of the most popular podcasts on its platform.
CTO and co-founder Acast Johan Billgren told TNW that the algorithm can recommend relevant shows that are in completely different categories:
For example, the Another Mother Runner podcast that talks about running, parenthood, and work/life balance is categorized as ‘fitness and health’ based on the words that are visible and trackable. But Acast can generate recommendations based on Another Mother Runner for its listeners of shows that are actually about work/life balance, parenthood, running, without the show ever being tagged or categorized that way.
Acast began testing the feature last October and found that users are 49 percent more likely to listen to multiple episodes of a show recommended by the new system.
That sounds useful – I’m constantly hunting for new podcasts to sink my teeth into and am looking forward to Acast’s recommendations after it analyzes my mix of favorites that range from true crime to cooking to comedians’ talk shows and architecture.