Akamai: OTT video streaming quality experience drives viewer loyalty, provider success

Akamai's NOCC (network operations control center) in Cambridge, Mass. Image: Akamai
SVOD brands tend to lose the most engagement due to buffering, while TVOD models suffer the most negative impact to brand loyalty if delivering low-quality experiences.

Akamai says that OTT video users's online video behaviors will fluctuate if buffering occurs regardless if the service is free or is part of a paid subscription model.

In a new study Akamai conducted with third-party research firm Sensum, Akamai found viewers disengage with emotive storylines and react negatively to low-quality streaming incidents like buffering regardless of the brand or interest in the content.

Sensum, an emotion-based software solutions company, has developed its Sensum Insights platform to gather emotional responses from audiences and turns them into actionable insights to help brands hone their products, user experiences, content, retail and immersive experiences.

RELATED: Akamai unveils new mobile optimizations for video delivery

Akamai’s study with Sensum used a variety of testing procedures including sensory, implicit and explicit responses from more than 1,200 participants. Each of the tests adhered to the Video Quality Experts Group (VQEG) standards to ensure results could not be contaminated.

Shane Keats, director of global industry marketing, media and entertainment for Akamai, told FierceOnlineVideo the goal with working with Sensum was to understand how OTT consumers react to quality fluctuations on paid versus free content.

“We wanted to see if we could develop a methodology to do cuts against OTT business models,” Keats said. “Are there differences in the way that people pay for their content that affect how they view quality?”

According to Sensum’s research, negative emotions increased 16% while engagement decreased nearly 20% as a result of these poor experiences. The survey also demonstrated that 76% of participants said they would stop using a service if issues such as buffering occurred several times.

Additionally, Akamai also created three fictitious brands to remove any previous emotional association with the business models and used the same content across all the brands to nullify the impact of content type on the respondents.

These brands included: 24x7 TV, Free TV and Quick TV. The 24x7 TV service was Akamai’s proxy for subscription services, while Quick TV was the transactional model and Free TV was for an ad-supported service.

Although each of these models offered consumers an all you consume access to content, the difference was how they paid for it. Users would have the option to watch content with skippable advertising, pay a monthly fee or pay per title.

“You would show these options before each user would watch content and that way they’re associating their viewing,” Keats said.

The research unveiled key trends about subscription video-on-demand (SVOD) and transactional video-on-demand (TVOD).

SVOD brands lose the most engagement due to buffering, while TVOD models suffer the most negative impact to brand loyalty if delivering low-quality experiences.

“With the AVOD and TVOD users there was not any big delta between how they react with the skin monitor, but SVOD were 175% more engaged during this peak sequence during the buffer event,” Keats said. “People who believe they are paying for something every month really love the high quality.”

Keats added that “subscription-based viewers are more responsive to higher quality streams.”

High-resolution video content with emotive storylines improve viewer engagement by more than 10%.

Interestingly, the study found that when buffering begins, viewers experience various emotional fluctuations:

Happiness drops 14%

Negative emotions (disgust and sadness) increase by an average of 8%

Viewers’ feeling of surprise increases 27%

Attention drops by 3% and focus decreases by 8%.

To measure these metrics, Sensum used facial coding software. People who were online had to have their laptop camera on and the camera captured users’ facial twitches while they were consuming video.

“This is like the guy who you’re gambling with in a poker game where you say the other play has got what is known as a ‘tell,’” Keats said. “There’s something about the other player’s eyebrow quivers or something about the way their nose twitches and that’s a ‘tell’ that a good poker player can use to understand what that guy is really feeling.”