'

Buzzfeed Content Teardown

Понравилась презентация – покажи это...





Слайд 0

Content Teardown: Buzzfeed Facebook Page hello@decisive.is 1


Слайд 1

Methodology We generate a set of facts about each post and use them to build a predictive model of the Decisive Content Score. We use various techniques from robust statistics to build these multiple regressions with each fact encoded as a regression variable. The following slides include a few of our actionable findings. 2


Слайд 2

What’s In The Decisive Score? The Decisive Score takes into account traditional engagement metrics such as reach, likes, favorites, comments, but also measures how willing your audience is to share your content, and the emotions your audience express. The Decisive Score gives you a comprehensive view of how your content is performing across accounts and platforms. Likes/Favorites Low Impact Sentiment High Impact Shares High Impact Mentions High Impact Comments 3 Medium Impact


Слайд 3

Emojis Can Predict Performance Our analysis shows the types of emojis that users respond with on a Buzzfeed post is a strong predictor of how well a post will perform. When


Слайд 4

Squares Work Best Across Buzzfeed's accounts, square images boosts the Decisive Score. Square images work so well because they are optimized for mobile screens. # of Posts by Image Size 500px 3,000 +2.5 Points Higher +14.9 Points Higher 2,250 1,500 2,852 500px 750 1,407 0 Wide 5 Tall 1,114 Square


Слайд 5

Video Is WAY more engaging # of Posts by Content Type 5,000 3,750 2,500 4,723 + 22.8 Points Higher 1,250 0 6 649 Photo Video


Слайд 6

Hashtags also drive higher engagement! # of Posts 6,000 4,500 +8 Points Higher 3,000 5,304 1,500 0 7 69 Without Hashtag With


Слайд 7

Top 10 Response Keywords By and large people come to Buzzed to be entertained. Rank Keyword lol 8388 people 4555 love 3891 cute 2075 haha 2053 buzzfeed 1927 time 1656 make 1401 funny 1344 hahaha 1306 8


Слайд 8

Other Interesting Findings Content posted after 9pm had moderately better scores than other times. Lighter images had moderately better scores than darker ones. Images without a caption had significantly lower scores. 9


Слайд 9

Send us an email to say hi


Слайд 10


×

HTML:





Ссылка: