Axios-NewsWhip 2020 attention tracker: Bloomberg finally upends the national conversation
Stories about Michael Bloomberg last week generated 9.4 million interactions on social media — more than twice his previous high. Still, he's getting lapped by Bernie Sanders, according to data from NewsWhip provided exclusively to Axios.
Why it matters: This was the point when Bloomberg converted massive spending into significant organic interest in his campaign, but it may be no match for the Sanders grassroots army.
The big picture: Since joining the race, Bloomberg has spent his way into the national conversation, while lagging behind in organic interest.
- But as he has climbed in the polls, he has entered the crosshairs of the national conversation — and may now be establishment Democrats' best chance at keeping the nomination away from Sanders.
Yes, but: Sanders is also moving into a new tier. His 18.5 million interactions last week were higher than any Democratic candidate has generated throughout the entire election cycle.
- Over the last four weeks, his 50 million interactions are 20 million more than Biden, who has the second most.
- Bloomberg's team fears that if competing moderates don't drop out, Sanders' delegate lead could become insurmountable.
By the numbers: While still far off from Sanders, Bloomberg is surging into second place by many measures.
- In less than a month, he's erased a 22.5-point deficit and now trails by just 1.7 points in the RealClearPolitics polling average.
- He surpassed Biden in social media interactions this week. His 9.4 million interactions were more than any candidate other than Biden or Sanders have garnered in a single week all cycle.
- Over the last week, he has eclipsed Joe Biden for the second-most cable news and nightly network news mentions, according to the the Internet Archive Television News Archive.
Between the lines: Bloomberg's surging profile has been boosted by the volume of coverage about him.
- In the first six weeks of his campaign from early November to mid-December — when he was polling in the low single digits — Bloomberg had 24,754 articles written about him. That compared to 17,259 for Warren, 11,954 for Sanders, 5,242 for Buttigieg and 2,243 for Klobuchar.
Our 2020 attention tracker is based on data from NewsWhip exclusively provided to Axios as part of a project that will regularly update throughout the 2020 campaign.
Why this tracker matters: The data on interactions — including likes, comments and shares — highlights an important, but under-appreciated element of an election: the ability to see beyond our own social feeds and understand the broader universe playing out of how candidates and issues are moving the minds of voters.
- It measures enthusiasm in a way that traditional polling does not.
- The sample size taken from these social media platforms is massive.
- Social media is powered by emotion-driven content, and emotional responses are likely to be aligned with a voter's true beliefs in a way that can be masked in polling.
While the volume of interactions does not gauge the sentiment of the reactions, the ability to generate reach allows a candidate to expand the universe of potential voters.
- Bots also cannot be ignored, and we will point out in this space if there are documented instances of bot activity for certain candidates or issues.
Methodology: This project measures the number of social media interactions generated on stories published about the 2020 candidates and issues.
- Interactions are calculated from reactions, comments and shares on those stories on Facebook as well as the number of shares from more than 300,000 influential Twitter accounts and retweets and likes on those posts.
- Tracked published stories come from a defined universe of more than 450,000 domains.
- A story registers for a candidate or issue if the keyword is mentioned in the headline, summary or URL of the story.
- Our search format for candidates looks like: "Joe Biden" OR ("Biden" AND ("President" OR "2020" OR "election" OR "Democrats" OR "primary")).
- For issues, we use a keyword tree for related terms.