Twitter hosted 600 hours of live video content programming on the platform during Q4 2016, Twitter COO Anthony Noto said during a recent analyst call. Out of this, about 50% of live content was in sports, around 40% in news and politics and 10% in live content created by users. Noto added that Twitter’s live video streaming deal with NFL contributed significantly to live video programming.

About 50% of viewers engaged with NFL scores and video related content were younger than 25 years old, around 25% of the overall audience came outside the USA, and 12% of the engaged users were non-Twitter (or non-logged in users). The company also hinted that it plans to renew the partnership with NFL for a longer time-period. “All those things (growth in live video) rang true and we will look to partner with them (NFL) in a bigger way,” Noto added.

On advertising

“The majority of our revenue remains branded advertising. It’s the area that we’ve had the most success,” Twitter CEO Jack Dorsey said during the call. Twitter’s branded ads are offered via “video products, especially the Amplify product and the live product”. Dorsey added that the company plans on investing more into such ad products. Apart from video-based ads, Twitter also offers dynamic ads—ads that are relevant to a user—on mobile web, app and desktops. “Those ads had 95% completion rates with sound on,” Noto said.

Focus on relevant content and machine learning

Dorsey mentioned that Twitter will continue to tweak and update its timeline, notifications, and especially on ‘Explore’ tab which aggregates relevant search results, moments, trending tweets and topic. “We’re going to put a lot of attention into Explore, and making sure that we show more topics faster. This is also an area where we can show off our live streaming games, debates, elections, and events,” added Dorsey. Note that Instagram already has an ‘Explore Tab’ which aggregates relevant stories, videos, pictures, etc.

However, Dorsey clarified that Twitter’s “fundamental” use case still remains on the ability “to show what’s happening immediately, being able to get people their news immediately, about anything that’s new and noteworthy, and also to show all the commentary and conversations around it.” Dorsey also said that Twitter will invest more in machine learning and artificial intelligence to lead users to relevant content. “…we’re looking at a lot of opportunities to organize all the tweets around relevance but also around the topics and the interest and the passions that people care about,” he added.

User metrics
– Twitter currently has over 319 million monthly active users (MAUs) globally, up 4% from 305 million MAUs in the same quarter last year. In the previous quarter, Twitter reported more than 317 million MAUs.
– There were 67 million MAUs in the US at the end of Q4 2016. MAUs in the US have remained almost flat throughout the whole of last year. The company had reported 65 million MAUs in the US for Q1 2016, 66 million MAUs reported in Q2 2015, and 67 million during Q3 2016 and Q4 2016.
– International MAUs stood at 252 million for the quarter, up 5% from 241 million MAUs in the same quarter last year. In the previous quarter, Twitter reported 250 million international MAUs.

Financials
– Twitter reported revenues of $717 million for Q4 2016, up marginally from $710 million revenues reported in Q4 2015. In the previous quarter, Twitter posted overall revenues of $616 million.
– The company’s net losses were $167 million during this quarter, compared to the loss of $90 million in the same quarter last year (Q4 2015). In the previous quarter (Q3 2016), Twitter posted net losses of $103 million.
– Advertising revenue accounted for 89% or $638 million of the revenues, while data licensing & other accounted for the remaining $79 million revenues.
– Advertising revenue grew 17% QoQ from $545 million in the previous quarter.
– EBITDA stood at $215 million for the quarter, up 12% YoY from $191 million in Q4 2015. In the previous quarter, EBITDA stood at $181 million.

 Download: Press Release | Presentation | Transcript