TiVo: Viewers use an average of 9 streaming services

With the streaming industry becoming increasingly saturated, making sense of viewer data and consumption habits is key. According to TiVo’s most recent video trends survey, consumers' average number of streaming services peaked to about nine apps in Q4 2021.

And Chris Ambrozic, SVP of product management for TiVo Discovery Solutions, hinted to Fierce Video the average number of apps continues to tick up in TiVo’s upcoming consumer survey – scheduled for release in September.

Ambrozic noted that uptick is being fueled by AVOD and free ad-supported streaming TV (FAST) services. U.S. homes have increased their AVOD service adoption, with a 29% increase in homes streaming AVODs so far this year, according to Comscore.

“People are adopting more AVOD apps, that’s where the growth seems to be coming from in this most recent survey,” he said. The results highlighted that out of the average user’s nine apps, about 2.4 of those streaming services were designated as free.

Similarly, recent survey data from Hub Entertainment Research showed 69% of respondents saying it takes eight or more services meet their TV lineup needs “very well.”

Despite frequent conversation about the future of aggregation for OTT platforms, viewers continue to build up their streaming service collections. Ambrozic thinks what’s happening is people are churning out of apps at a fast rate, depending on what content is available and how much they are willing to spend.

“They come in, they consume some content and then they jump out and go the next [app],” he said. “Certainly the industry has done a very good job making a lot of really good high quality content that is attractive to people."

How viewers seek content

As those viewers churn from their subscription-based services, they’re adding more free services. Navigating the aggregation landscape, Ambrozic noted, becomes more complicated if a show jumps around from one service to the next.

“As industry people, we follow that pretty easily,” he said. “But we can’t ask the consumers to do that – it’s confusing to them.”

Just as shows and movies jump around over the streaming landscape, so are viewers. Ambrozic referenced an IGN study in which 62% of respondents said they canceled a particular service after finishing a TV series or movie.

So people are coming into an app with specific content in mind, he said, suggesting viewers aren’t necessarily indecisive about what they want to watch. But that doesn’t mean people are constantly looking for new titles.

“We have some other statistics that show people on the weekdays have a lot of what I’ll call cyclical behavior,” Ambrozic continued. “So they’re not always jumping into a new show or a new movie. During the weekdays, they’re doing a little bit more like getting back into the show they had been bingeing.”

Whereas on the weekends, consumers are more open to experimentation. And that makes sense, he pointed out, as that’s when people generally have the most leisure time.

“Sometimes people have a very high degree of intent. They know exactly what they’re looking for and they will just watch that and bounce [to another service],” Ambrozic said. “Other times they’re more what we call suggestive in nature. You know, ‘suggest to me what might be good.’”

To give viewers better suggestions, streamers are placing importance on user experience. Amazon Prime, for instance, is undergoing a massive overhaul of its UI, incorporating features like a Top 10 Chart of content on its home page.

Suggestive moments are critical, Ambrozic points out, and those moments tend to occur at the end of a show or season.

“You’ve got to have the right thing for them at that time or they’re out,” he said. “There’s so much out there that’s trying to grab our attention and keep our attention, that you have to be really good at that moment when somebody picks a new show.”

Personalization data

He went on to say the methods for user personalization and recommendations have become more complex – algorithm-wise – than how they were about a decade ago. Nowadays, dozens of algorithms are vying to understand and respond to different user behaviors.

Algorithms also play a part in a user’s Match Score on TiVo’s Stream 4K product. The Match Score offers personalized ratings of each streaming service on a scale of 1 to 100.

Personalization typically uses implicit and explicit signals, Ambrozic explained. Implicit signals track, for instance, what a viewer watches and for how long. Whereas an explicit signal is how the viewer reacts to content, whether it’s via an action like a thumbs-up or a star rating system.

“We use those same signals for the TiVo Match Score. But we use them in a slightly different way,” he said. While a personalization engine provides users more generic recommendations – regardless of where the content is coming from – the TiVo Match Score analyzes content on a catalog basis to determine suggestions for the viewer.

A streaming service with a Match Score of 80 has more content in line with what the targeted user wants to watch, Ambrozic noted, as opposed to a service with a score of 60.

“If you understand your target market well, you should hopefully have the content for them and get that in front of them,” he said. “So that you have a high score, so that they stay with you.”

An April MoffettNathanson report pointed out cord cutters are increasingly switching to streaming services because of content, rather than streaming’s comparative cost to cable.

Value of voice

Voice control is another layer of personalization, helping people locate content. TiVo’s survey indicated 33% of respondents use voice commands with at least one of their video entertainment devices.

An effective voice system is one that differentiates between users of the same household, Ambrozic said. Algorithms can take an audio file, which is then translated into a text string. But with that technology comes the topic of data privacy, he added.

“If you’re going to do this in Europe, where you have to be adherent to GDPR regulation…that voice imprint, if you will, is considered personally identifiable information,” said Ambrozic. “So you’ve got to be respectful of that.”

If streaming services can manage to make voice control systems more efficient, viewers stand to further their engagement with those platforms. Ambrozic pointed out while the survey says voice control varies between regions – 48% usage in the West Coast vs. 29% in the Midwest – there’s a general upward trend towards personalized voice systems.

“You can think of your voice and the engagement with the discovery system – the system that’s trying to find you content – as like a wormhole to any destination in the universe,” he said.