Mass Media

A few weeks ago, we conducted an online survey to better understand our audience’s opinions and general knowledge of the regulation of social media websites. Our overall goal with this survey was to determine whether this issue mattered to social media users. Our overall finding was that out of the 30 users surveyed, the majority of them are aware that social media websites sell their information to advertisers.

The majority of people surveyed fell between the ages of 13 to 25, which is the same audience targeted by advertisers via social media websites. Over 60 per cent of users were registered with more than two websites. 

When asked if they felt that their privacy had been invaded by these websites, over 40 per cent agreed with this statement while 30 per cent said they felt somewhat invaded.  Over 60 per cent were aware that there is no current law regulating social media websites. Moreover, an astonishing 80 per cent agreed that this is in fact and issue and it matters to them.

When asked who should be responsible for ensuring users protection online, 30.4 per cent said themselves (social media users) with another 30.4 per cent stating that the government, websites, marketers and users should join forces into regulating social media. What we found surprising was that 20 per cent of users felt this was the government’s responsibility.

Over 60 per cent stated that they have never been influenced by such advertisements on social media websites. If this is the case, why are more and more advertisers reaching out to these users with their ads?

While social media and video game companies have been surprisingly successful at hiding their use of persuasive design from the public, one breakthrough occurred in 2017 when Facebook documents were leaked to The Australian. The internal report crafted by Facebook executives showed the social network boasting to advertisers that by monitoring posts, interactions, and photos in real time, the network is able to track when teens feel “insecure,” “worthless,” “stressed,” “useless” and a “failure.” Why would the social network do this? The report also bragged about Facebook’s ability to micro-target ads down to “moments when young people need a confidence boost.”

By Joey Weed and Brandon Carter

The summer interns for both the Social Media Desk and Audience Insights have departed. Their parting gift was this analysis. – Wright

Almost every weekday since May, NPR has published a Facebook Live feature called Cereal. Hosted by NPR’s Two-Way bloggers, each episode covers the morning’s top headlines and then takes a deep-dive into the most interesting story of the day. Now more than three months into this experiment, how’s it doing?

We focused our analysis on how viewers watched Cereal in their news feeds, how well each episode retained viewers and how the series has been performing over time.

We looked at 54 episodes, running from May through July. The data gathered from Facebook Insights allowed us to look at video views for 3 seconds (Total Views), 10 seconds and 30 seconds. Facebook also measures how many people are watching an episode – whether it’s autoplaying in their newsfeed or whether they clicked directly on the video. Finally, we can see how many viewers were watching at particular spots during each live broadcast to better understand retention.

How has Cereal been doing over time?

At the start of the series, we saw a higher number of negative interactions with the audience. We measured these by adding up the negative signals for each video (hiding the video, hiding all videos from the NPR page or un-liking the NPR page) and dividing that number by post impressions. This new variable allowed us to compare the frequency of negative interactions for videos with varying degrees of impressions.

The negative interaction trend we observed suggests that NPR’s Facebook audience is becoming more accustomed to seeing Cereal in their news feeds, or that they are no longer seeing the notifications that bothered them in the first place. As both negative interactions and 30-second views level off, we believe Cereal has settled into a more-or-less consistent audience pattern.

Are viewers watching Cereal with sound?

We found that 35% of views are played with sound during a typical episode. We saw a moderate correlation between videos accompanied by shorter text descriptions and increases in views with the sound turned on. This insight reveals that the video description could have an impact on how people engage with the video.

How long is our audience watching Cereal?

The average Cereal viewer watches approximately 30 seconds of any single episode. This number has remained stable from month to month.

While average time spent watching Cereal videos has remained constant over time, the average length of our episodes increased from 17 minutes in May to 22 minutes in July. The result is that the average user is seeing a smaller percentage of the overall video (see chart below).

Future areas for analysis.

Two areas we’d like to look at in the future are the performance of breaking news and the performance of streams with visual aids or props. Initial analysis shows that segments with more to look at can boost retention compared to verbal descriptions of complex topics.

Cereal episodes have seen better retention when the breaking news topic is covered for the entire episode. Episodes that included other topics along with the breaking news of the moment did not fare as well.

As the samples sizes are small, we’ll reserve these topics for future analysis as the data set grows.

- Joey

Editorial takeaways from Brandon Carter:

Video description text matters. Our analysis shows that there’s some correlation between shorter descriptions and an increased number of views with the sound on. This finding highlights the importance the video description field for Facebook Live videos. Like the lead of a story, the most important names and information should appear at the beginning of the description. We should keep that text short, sweet and to the point. This will quickly inform viewers of what they’ll see and, hopefully, win their interest.

Video length: Shorter can be better. The average length of Cereal episodes has increased over time, while the amount of time the average viewer spends watching Cereal has stayed the same. We want our viewers to watch more of each episode. One strategy for increasing this number could be the shortening of episodes in conjunction with a greater focus on deep dives. 

Trusting our news judgement is key. Although we don’t have enough data for a firm conclusion, breaking news seems to bring spikes in viewership and viewer retention. While breaking news doesn’t happen every day, we should continue to embrace “news of the now” for the same reason. A great example of this was our August 3 episode, in which Two-Way blogger Eyder Peralta chatted with politics editor Domenico Montanaro about Donald Trump. While none of the topics discussed were big enough news to be considered “breaking,” the overall conversation was fully on the news. We’ll have to watch the numbers to see if they confirm our hunch that both “being on the news” and “breaking news” bring in and retain audience.

Keep it visual. Our analysis suggested that episodes with additional visual element – Skype calls, diagrams, maps, etc. – may increase retention. We think it would make sense for our producers to work alongside the Two-Way team to bring a greater diversity of visual elements to each episode.