how not to measure twitter influence

Post on 10-May-2015

385 Views

Category:

Technology

2 Downloads

Preview:

Click to see full reader

DESCRIPTION

These are the slides of a keynote I gave at Emerce Eday on 25 October 2012 in Rotterdam. The short description of my talk was as follows: With the ongoing rise of third party applications like Klout, tools for measuring Twitter influence are important to understand. This presentation takes a look at the different ways in which influence measures have been developed for Twitter. In particular it will use the case study of the UK riots of 2011 for which a database of 2.6 million tweets was collected in collaboration with Twitter and The Guardian newspaper. By examining the top 1000 most tweeted accounts, it will give further insight in how influence worked during this crisis event, specifically highlighting the emergence of the ‘ordinary influential’ during 2011 as well as how large organisations have incorporated social media practices.

TRANSCRIPT

How (not) to measure Twitter influence

Farida VisInformation School, University of

Sheffield

@flygirltwo

REAL TIME

BIG DATA

MOBILE

TWITTER

FACEBOOK

[NAME OF NEXT BIG SNS HERE]

GEOLOCATION

STREAMING

INFLUENCE

KLOUT

PEERINDEX

Following Rules and Best Practice – Twitter Help Center

We don’t limit the number of followers you can have. However, we do monitor how aggressively users follow other users. We try to make sure that none of our limits restrain reasonable usage, and will not affect most Twitter users.

Aggressive following is define as indiscriminate following a few users if their accounts seem interesting is normal and is not considered aggressive.

… every user can follow 2000 people total. Once you’ve 2000 users, there are limits to the number of additional users you can follow: this limit is different for every user and is based on you ratio of followers to following.

‘A score in the 40s suggests a strong, but niche, following’

‘… influence is about engagement and motivation, not just racking up legions of followers’ (New York Times)

Sleep deprivation…

Facebook

Community

Not on KLOUT, still scored

THE ALGORITHM

Aaron Zinman

MIT SociableMedia Group

Illegal and legalEducationSportOnline

UK TWITTERATI

2011 UK Twitterati (The Independent)

1. Sarah Brown – Campaigner (PI: 93)

2. Richard Bacon – Broadcaster (PI: 92)

3. Eddie Izzard – Comedian (PI: 89)

2012 UK Twitterati (The Independent)

1. Richard Branson – Tycoon (PI: 93)

2. Sarah Brown – Campaigner (PI: 92)

=2. Alan Carr, Broadcaster (PI: 92)

Tweets as: @richardbranson

PeerIndex:93

Authority: 95

Activity: 46

Audience: 97

Follows: 6,414

Followers: 1,834,516

NO Jon Hickman on PeerIndex

No score/profile created for him

Free stuff

More than 2,500

companies using

Klout’s data

(NYT, 26.6.2011)

“For the first time, we’re all on an even playing field”

Joe Fernandez

Chief executive and co founder of Klout

BIG DATA

NOT ENOUGH

GIANT DATA

More than 12 billion signals a day into a Hive data warehouse of more than 1

trillion rows

Hundreds of millions of user profiles

Klout blog 11.10.2012

2.6 million riot tweets

Lisa Evans, 8.12.2011

± 45% Mainstream media

Viral

Image

Influence

Mentions ≠ insight

Online influence ≠

Offline influence

BIEBER VS OBAMA

2012 UK Independent voices (The Independent)

UK NEWS: Owen Jones –

writer / researcher (PI: 65)

WORLD NEWS: Hossam el-Hamalawy –

Egyptian journalist/dissident (PI: 64)

PEOPLE: Wael Ghonim –

Egyptian Google staffer (PI: 41)

2011 rise of the ordinary influential?

CONTEXT

CONTEXT

CONTEXT

Small data = meaning

Gaming the system

Implications of reducing

people to a number

top related