This presentation shows how to use social media data and analysis in academic research in the field of marketing, economics, consumer behavior, and psychology. LiveWall (www.li.vewall.com) collects social data and allows you to use it for research.
2. LiveWall
• Founded in 2011 by former Tilburg University students with
years of experience in IT and media
• Part of IntoApps and ServingSocial
• Goal: to connect online and offline audiences & and to make
social media visible & concrete
3. Customers
We have a wide range of hundreds of users and customers,
including brands, service providers, sports, event companies, B2B
4. LiveWall platform
• Content aggregation platform for social media sources
Data from Twitter, Facebook, Instagram, Foursquare, Youtube,
Flickr
• Display messages on screens
For events, customer service, internal usage
• Statistics of social media usage
Metrics, text analysis etc.
16. At social media we talk about..
how we work
what we know
how we learn
17. At social media we talk about..
how we work
what we know
how we learn
who we like
18. At social media we talk about..
how we work
what we know
how we learn
who we like
what interests us
19. At social media we talk about..
how we work
what we know what we hate
how we learn
who we like
what interests us
20. At social media we talk about..
how we work
what we know what we hate
how we learn how we evaluate
who we like
what interests us
21. At social media we talk about..
how we work
what we know what we hate
how we learn how we evaluate
who we like our emotions
what interests us
22. At social media we talk about..
how we work
what we know what we hate
how we learn how we evaluate
who we like our emotions
what interests us our life
23. ..and social media captures this social
interaction digitally in huge databases
24. ..and social media captures this social
interaction digitally in huge databases
Why not use this in research in the field of
communication, marketing, economics and psychology?
25. Value starting to get noticed
The Wall street Journal
The Guardian The Boston Globe
27. Getting data
• Social media platforms give access to their data
– Through the use of API’s (Application Interfaces)
– For external software
– Easily download data and use it
– In standardized formats per platform
• LiveWall’s platform can load that data and process all data in a
standard or custom format
28. Twitter: terminology
• Follow: subscribe to see the posts of another user
• Follower: someone who follows you
• Tweet: one post of 140 characters
• Hashtag: Word with # in front to categorize the post
– People can follow one subject
– Free to choose
• Retweet: resend the message of someone else
– Spread news
• Trending topic: the most popular subjects
• Mention: a tweet that addresses (mentions) another user in the post (with @ sign)
29. Twitter: available data
• Tweets list
– Tweet text and details based upon a query by hashtag, language, username or location
– Number of tweets per hashtag
• Individual tweet
– Retweets
– Assigned media: video/picture
– Source: web/mobile
• User
– Number of followers, friends and favourites
– Location data (limited)
– Mentions
– Description
30. Facebook: terminology
• Post / status: one content element that can include link, image, video or just text
• Like: someone gives positive feedback on a post, or says it wants to connect to
the page or content
• Timeline: the personal collection of photos, messages and experience of every
user
• (Fan)Page: page of a brand, company or event that includes posts, video’s,
pictures and apps of a brand. On the fanpage both the holder of the page and the
audience can communicate
• Fan: Someone who likes your fanpage and thereby subscribes to receive all the
updates that are placed at the Facebook page on their personal timeline
31. Facebook: available data
• Fanpage
– All posts
– Likes and comments
– Total number of fans
– Photo’s, video’s
• Fanpage (after approval of page administrator)
– Impressions (view of any content), by source
– Engagement: total of click, like respond etc.
– Consumptions: people who clicked contnet
– Fans by city, country, gender, age
– Page views
32. Facebook: available data
• Post (after approval of page administrator)
– Impressions
– Engagement
– Consumptions
– Negative feedback
• User (after approval of user)
– Birthday
– Hometown
– Work
– Education
– Languages
– All friends
– All liked pages and subjects
– Location based checkins
– All posts including text
– Photo’s, video’s
33. What can LiveWall do
• Collect metrics over time, E.g.
– Hourly number of tweets per channel (Twitter, Instagram etc.)
– Number of retweets per post
– Impressions and engagement of Facebook page per day
– Amount of social media posts by the company itself per day
Etc.
• Text analysis
– Associations
– Sentiment
– Audience interests
• Set up a custom application or construction for a research
38. Examples
• Is there more feedback on negative messages? (real data,
experimental)
– Sentiment and retweets and/or reach in variance analysis
• What type of message has the best effect
– Text / image vs shares/likes/retweets
– Second person / first person vs shares/likes/retweets
• Can social media feedback predict share value
– Regression with share value over time, number of posts over time,
post sentiment
39. Examples
• Which brands clusters are there in industry X
– Cluster analysis of associations of brands (Red Bull – Vodka)
• What types of brands tend to have the biggest social media success
– Compare brand groups based on followers, likes and posts about the
brand
• Do you have the same social relationships online as you have offline?
– Compare interactions with friends online with the top rated friends
offline
40. Examples
• What is the difference between social content on Facebook in
comparison to Twitter
– Compare Twitter and Facebook accounts of one organisation and
see how the reactions differ
• Does peer endorsement work better than celebrity
endorsement?
– One part of the group sees Facebook friends that endorse a
brand, the other part of the group sees celebrities to endorse the
brand
41. Examples
• What are the associations people make with certain
product groups (for example sigarets )?
- Input can be used for designing an effective campaign for or
against usage of the product.
• When a brand is associated in social media with a certain word
like ‘amazing’ or ‘splendid’, does mentioning the word in
campaigns or commercials lead to a higher liking?
43. Web
www.intoapps.nl
www.livewall.nl
Where
Online?
Eelco
eelco@livewall.nl
@bizzsms
Notas do Editor
Can we apply swarm theory to social data analysis? As individuals, we might not be able to hold onto or understand a dataset, but as a group, we can come at a dataset from different perspectives, look at very small parts, and then as an end result -- extract real, worthwhile meaning.