Ch 1.
SOCIAL MEDIA
ANALYTICS: AN OVERVIEW
--- Sakshi Surve
Contents
● Types of Social Media
● Characteristics of Social Media
● Purpose of Social Media Analytics Social Media vs. Traditional Business Analytics
● Seven Layers of Social Media Analytics
○ Layer One: Text
○ Layer Two: Networks
○ Layer Three: Actions
○ Layer Four: Mobile
○ Layer Five: Hyperlinks
○ Layer Six: Location
○ Layer Seven: Search Engines
● Types of Social Media Analytics
○ Descriptive Analytics
○ Predictive Analytics
○ Prescriptive Analytics
● Social Media Analytics Cycle
● Challenges to Social Media Analytics
Types Of Social Media :
● Based on Authentication or access mechanisms, social media tools are available in two forms:
○ Internet-based
○ Smartphone-based.
● Internet-based social media platforms are generally accessed through e-mail IDs.
Example : Facebook, LinkedIn, Cyworld, and Google+ are examples of Internet-based social media.
● Smartphone-based social media platforms are accessed through mobile phone numbers; that is, users can
only log in using mobile phone numbers.
Example : KaKao Talk, Tender, and 1KM
Characteristics of Social Media :
● SOCIAL MEDIA IS MANY-TO-MANY
○ Social media enables interaction among the users in a many-to-many fashion. This is unlike conventional
technological media such as print, radio, telephone, and television.
● SOCIAL MEDIA IS PARTICIPATORY
○ Unlike conventional technologies, social media encourages participation and feedback from users. Social
media users can participate in online discourse through blogging, comments, tagging, and sharing content.
● SOCIAL MEDIA IS USER OWNED
○ While social media platforms are provided by corporations (such as Google and Facebook), the content is
generated, owned, and controlled by social media users. Without the user-generated contents and active
involvement from the users, social media would be empty, non existing
`
● SOCIAL MEDIA IS CONVERSATIONAL
○ Because of the many-to-many abilities, it allows the masses to communicate and collaborate in real time.
● SOCIAL MEDIA ENABLES OPENNESS
○ Social media provides new opportunities to access data and information through Web 2.0
channels.
● SOCIAL MEDIA IS RELATIONSHIP ORIENTED
○ Most social media tools allow users to easily establish and maintain social and professional
relationships and ties. Some social media tools, such as Facebook, are solely focused on
personal relationships, whereas others, such as Twitter, are focused on professional
relationships.
● SOCIAL MEDIA IS FREE AND EASY TO USE
○ Being free and easy to use are two of the reasons that social media has expanded in such a
rate.
Social Media Analytics: Overview
● Social media analytics is the process of extracting valuable insights from social media
data to perform decision making.
● By analyzing social media data, businesses can increase brand loyalty, generate leads,
drive traffic, and make forecasts.
● Social media analytics can also be used to increase awareness of a brand and drive
users to a website for the latest news and information.
● Social media growth, tools, and big data also present huge opportunity to market
products, enhance brand loyalty, network with customers, crowd-source ideas, drive
sales, and mine business insights.
Purpose of Social Media Analytics
● The purpose of social media analytics is to enable informed decision making by
leveraging social media data. This involves answering questions such as:
○ What are customers saying about a brand or product on social media?
○ Which content posted on social media is resonating more with customers?
○ How can social media data be used to improve products or services?
○ Is the social media conversation about a company, product, or service positive, negative, or
neutral?
○ How can social media be used to promote brand awareness?
○ Who are the influential followers, fans, and friends on social media?
○ Who are the influential nodes (people and organizations) on social media and their position in
the network?
○ Which social media platforms are driving the most traffic to a corporate website?
○ Where are the social media customers located geographically?
○ Which keywords and terms are trending on social media?
○ How active is social media in a business and how many people are connected with the
company
○ How are competitors performing on social media?
Social Media Vs Traditional Business Analytics
● The main difference between social media analytics and traditional business analytics is the
source, type, and nature of the data being mined.
● Social media data is diverse, high volume, real-time, and stored in third-party databases in
semistructured and unstructured formats, while traditional business data is mostly stored in
databases and spreadsheets in machine-readable i.e. Structured format.
● Social media data is socialized in nature and originates from the public internet, while traditional
business data is bureaucratic and formal in nature and is controlled by organizations.
● The value of social media data is determined by the extent to which it is shared with other social
entities, while the value of traditional business data is often confined within organizational
databases and serves as a source of competitive advantage.
Types of Data :
Social Media Vs Traditional Business Analytics
Social Media Vs Traditional Business Analytics
Seven Layers of Social Media Analytics
● Social media has a minimum seven layers of data
● Each layer carries potentially valuable information and insights that can be
harvested for business intelligence purposes.
● Out of the seven layers, some are visible or easily identifiable (e.g., text and
actions) and other are invisible (e.g., social media and hyperlink networks).
● The following are seven social media layers
1. Text
2. Networks
3. Actions
4. Hyperlinks
5. Mobile
6. Location
7. Search engines
Seven Layers of Social Media Analytics
All Are
Interconnected !!!!
Example : A Restaurant's Social Media Marketing
Campaign
● The restaurant posts a description of a new dish across its social media platforms, including
appealing visuals and hashtags to attract attention. "Introducing our signature Spicy Thai
Noodles—bursting with authentic flavors! #FoodieHeaven“ ….(Text, image….Multimedia)
● Geotargeted ads ensure the post appears to users within 10 miles of the
restaurant…..(Location)
● The post reaches a broader audience ..A customer tags friends on Instagram to check out the
dish…..(Network)
● A search for "Spicy Thai Noodles near me" leads the customer to the restaurant’s website and
reviews….. (Search Engine)
● The post contains a clickable link leading to the restaurant’s website, menu, or reservation
system. Reserve your table now: [Link]/reservations... (Hyperlink)
● A customer browsing Instagram clicks the link to book a table during their lunch break….
(Mobile)
● After a good experience, Users comment, "Looks delicious! What’s the price?" or share it on
their own feeds….(Action)
See how it works …..
User posts TEXT →
Network spreads it →
Actions increase reach →
Hyperlink drives traffic →
Mobile enables instant posting →
Location attracts nearby users →
Search engines boost visibility
1. Text :
○ TEXT is one of the fundamental elements of the social media platforms.
○ Textual elements of social media include comments, tweets, blog posts, product reviews, and
status updates.
○ Social media text analytics, also known as text mining, is a technique to extract, analyze, and
interpret hidden business insights from textual elements of social media content.
○ Organizations use text analysis techniques to extract hidden valuable meaning, patterns, and
structures from the user-generated social media text for business intelligence purposes.
○ It is used to understand user sentiments and identify emerging themes and topics.
○ For example, Text Analytics is useful in gaining a quick and accurate understanding of the
emotion and sentiment expressed over social media channels (e.g., tweets or Facebook
comments) related to a brand or a new product launch.
2. Network :
● This includes the connections between users and the relationships between them, such as
followers and friends on social media. It is used to identify influential nodes (people and
organizations) and their position in the network.
● A social network is a group of nodes and links …..
○ nodes can represent social entities such as people and organizations.
○ Links represent their relationships, such as friendship and trade relations.
● Twitter follow-following network is an example of an online social media network. In a Twitter follow-
following network, nodes are the Twitter users, and links among the nodes represents the follow-following
relationship (i.e., who is following whom) among the users.
● When applied to Computer Supported Collaborative Learning CSCL, SNA is used to help understand how
learners collaborate in terms of amount, frequency, and length, as well as the quality, topic, and strategies
of communication
● Social Network Analytics is used in recommender systems development
Role of SNA in Recommendation Systems:
● By analyzing social interactions, relationships, and user behavior patterns, SNA helps in
building more accurate, personalized, and effective recommendation models.
● Examples :
○ If many of your friends watch and like a particular show, platforms like Netflix or Spotify
may recommend it to you.
○ Amazon’s "People who bought this also bought…" recommendations can incorporate
social data.
○ YouTube may recommend videos that are being actively watched and shared by a user's
network.
○ In Twitter, if an influencer tweets about a product, their followers are more likely to see
recommendations or ads for that product.
○ eBay or Amazon may recommend sellers/products highly rated by people you follow.
○ LinkedIn’s "People you may know" feature relies on this.
○ Spotify creating a "Group Playlist" based on shared preferences of connected users.
○ If a user likes a fitness product on Instagram, they might see ads for similar products on
Facebook.
3. Actions :
1. Actions: This includes the actions taken by users on social media, such as likes,
comments, shares, views, clicks and other interactions.
2. It is used to measure popularity, influence, and prediction in social media.
● Social media actions analytics deals with extraction, analysis, and interpretation of the insights contained in
the actions performed by social media users.
● Social media actions are of great value to social media marketers because of their role in increasing
revenue, brand value, and loyalty.
● Organizations can employ actions analytics to measure popularity and influence of a product, service, or
idea over social media.
○ For example, a brand marketer might be interested to know how popular their new product is among
social media users. Analyzing the Facebook likes and Twitter mentions, for example, can provide an
answer to the question.
4. Mobile :
● Mobile analytics deals with measuring and optimizing user engagement with mobile applications .
● This includes data related to the use of social media on mobile devices, such as the type of device used,
the operating system, and the location of the user
● Mobile analytics helps app developers and businesses monitor how users interact with their app, identify
issues, and continuously optimize features and performance to create a more engaging and successful
mobile experience.
● Mobile analytics refers to the process of collecting, measuring, and analyzing data related to user
interactions with mobile applications (Special-purpose software developed to perform certain tasks).
● It helps app developers and businesses understand how users engage with their apps, identify issues, and
make decisions to improve user experiences and optimize app performance.
Examples:
● A food delivery app may analyze user flows to see why users abandon their carts before placing an
order.
● If the app crashes frequently on a specific device, analytics tools help developers pinpoint and fix
the issue.
● A music streaming app like Spotify may recommend playlists based on the user’s listening history
and preferences.
● An e-learning app might send reminders to users who haven’t completed their courses.
● If most users access the app from iOS devices, the company might prioritize updates for iOS over
Android.
5. Hyperlinks :
● Hyperlinks are the pathways of social media traffic.
● Hyperlinks are references to web resources (such as a website, document, and files) that users can access
by clicking on it.
● Hyperlinks can link resources within a document (inter-linking) and among documents (intralinking). For
example, clicking on a hyperlink in a tweet can link you to other resources (e.g., websites) available over the
Internet
● This includes the links between social media posts and other websites, as well as the content of those
linked websites.
● Hyperlink analytics can also reveal, for example, Internet traffic patterns and sources of the incoming or
outgoing traffic to and from a website
Examples:
● A news website analyzes click-through rates (CTR) on article hyperlinks to determine which
headlines are most engaging.
● A fashion brand tracks the performance of links in its promotional emails to see which styles attract
the most attention.
● A domain ensures all research references link to authoritative sources.
● An online learning platform uses hyperlinks to link to related courses, so that user’s engagement is
increased.
● Track link clicks from social media posts to measure campaign success.
● Used to analyze the mentality of users, whether the links perform better in the header or the footer
of the website.
6. Location :
● This includes data on the geographical location of users and the location-
specific content they engage with on social media.
● It is used to mine and map the locations of social media users, content, and
data.
● Applications :
○ Geo Map related apps
○ Creation of bus services like CitiFlo
○ Fitness related apps
7. Search Engines :
● This includes data on how users find and access social media content
through search engines, as well as the keywords they use in their searches.
● It is used to analyze historical search data for trends analysis, and keyword
monitoring
Social Media Analytics Types
Any business analytics, social media analytics can take four forms to analyze, predict, and
optimize their strategies :
1) Descriptive Analytics (What)
2) Diagnostic Analysis (Why)
3) Predictive analytics (What is likely to happen)
4) Prescriptive analytics (How)
● Descriptive Analytics explains what happened.
● Diagnostic Analytics digs into why it so happened.
● Predictive Analytics estimates what is likely to happen next.
● Prescriptive Analytics provides actionable steps to influence future
outcomes.
Non technical example- Traffic Management in a city :
1. Descriptive: Identified a 30% increase in congestion. (Traffic sensors, GPS
data)
2. Diagnostic: Determined the causes to be construction, increased
vehicles, and poorly timed signals.
3. Predictive: Forecasted further congestion during upcoming events if no
action was taken.
4. Prescriptive: Implementing signal optimization, carpooling incentives,
and public transport discounts, leading to improved traffic conditions.
Traffic Management in a city (Considering Social Media):
1. Descriptive: Identified a 30% increase in congestion. (traffic-related
complaints, photos, or videos on platforms like Twitter or Facebook)
2. Diagnostic: Determined the causes to be construction, increased vehicles,
and poorly timed signals (trends and sentiments over social media).
3. Predictive: Forecasted further congestion during upcoming events if no
action was taken (tracking the future events announced over social media)
4. Prescriptive: Implementing signal optimization, carpooling incentives, and
public transport discounts, leading to improved traffic conditions (Engage with
citizens directly through social media platforms to share alternative routes,
real-time updates).
Example - Online Retail Store Analysis :
Let's consider an online retail store running analytics to improve its sales and
customer experience.
1. Descriptive Analytics (What happened?)
The retail store analyzes its sales data for the past month. The sales is found to
have dropped by 20% in the last month compared to the previous month
2. Diagnostic Analytics (Why did it happen?)
Sales declined due to low inventory of popular products and an increased
number of abandoned carts.
The team investigates the reasons behind the sales drop by Analyzing website
traffic, inventory records, and user behavior (e.g., heatmaps, cart data).
● Analyzed Reasons behind abandoned cards can be :
○ Unexpected Costs: High shipping fees, taxes, or additional charges at checkout.
○ Complicated Checkout Process: Lengthy or confusing forms to fill out.
○ Account Creation Requirement: Forcing users to create an account before purchasing.
○ Lack of Payment Options: Customers can't use their preferred payment method.
○ Security Concerns: Uncertainty about the website’s security for online payments.
○ Comparison Shopping: Customers add items to the cart but leave to compare prices
elsewhere.
○ Change of Mind: Customers reconsider their need for the items.
○ Technical Issues: Slow website loading, errors, or crashes.
3. Predictive Analytics (What is likely to happen next?)
Based on historical trends and external data, the store predicts future
outcomes.
"If inventory levels remain low and the checkout process isn't optimized, sales are likely to drop
further by an additional 15% next month.“
4. Prescriptive Analytics (How can we influence the future?)
The store implements actionable strategies to improve sales.
• Steps Taken:
• Increase inventory levels of popular products.
• Send personalized cart abandonment emails offering discounts.
• Simplify the checkout process
1. Descriptive analytics
● Descriptive analytics is the process of parsing historical data to better understand the changes
that have occurred in any situation
● Descriptive analytics can help to identify the areas of strength and weakness
● This involves gathering and describing social media data in the form of reports, visualizations,
and clustering to understand a business problem. Examples include actions analytics (e.g.,
number of likes, tweets, and views) and text analytics.
● Descriptive analytics answer the question “what is happening”? “What happened”? Comments
and posts can be grouped together, for instance, for the purpose of sentiment analysis
● Any time you gather a lot of similar data points in order to analyze them looking for patterns,
sentiments, and/or trends, you’re dealing with descriptive analytics.
2. Diagnostic analytics
● Diagnostic SMA analytics looks into the questions of “why something happened?”
● Diagnostic Analytics, as the name suggests, gives a diagnosis to a problem. It gives a detailed and in-
depth insight into the root cause of a problem.
● Enablers of diagnostics analytics include inferential statistics, behavioural analytics, correlations &
retrospective analysis and outcome being cause and effect analysis of a business issues.
● An example for diagnostic analytics can be an e-commerce company. Given the situation that the
sales of the company have gone down.
● The possible reasons behind this problem can be: the form didn't load correctly, the shipping
charges are high, and not enough payment methods are available.
● Taking the help of diagnostic analytics, the company comes out with a specific reason and then
works on that to resolve the issue.
3. Predictive analytics :
● Predictive analytics: This involves analyzing large amounts of accumulated social media data to
predict a future event.
○ For example, analyzing social media posts to predict future purchasing behavior or using
historical website visits to predict future sales figures.
● Predictive analytics uses models like data mining, AI, and machine learning to analyze current data
and forecast what might happen in specific scenarios.
● It can range from simple things, like predicting possible visits to a location based upon
posts expressing that intention, to forecasting entire trends and phenomena based
upon mentions.
● Social listening tools can help identify upcoming trends and shifts in consumer
behaviour by analysing large volumes of social media data and indicating the shifting
popularity of keywords.
4. Prescriptive analytics :
● Prescriptive analytics is a combination of data and various business rules. The data of
prescriptive analytics can be both internal (organizational inputs) and external (social
media insights).
● While predictive analytics help to predict the future, prescriptive analytics suggest the
best action to take when handling a scenario
● When combined with predictive analytics, it adds the benefit of manipulating a future
occurrence like mitigate future risk.
● Prescriptive analytics: This involves suggesting the best action to take when handling
a scenario based on analyzing patterns of behavior. It has not yet been widely applied
to social media data.
● Prescriptive analysis is the analysis of data with the intention of providing the best way
to proceed at any given moment.
● This can be applied to situations ranging from handling social media crises and
incidents (“how well this type of change will be accepted by our target audience?”) to
purchase preferences (“we’ve identified this group of customers – how do we optimize
our sales process to their habits?”).
Social Media Analytics Lifecycle
The social media analytics process involves six steps to mine desired business insights
from raw social media data.
The process begins with defining business goals and objectives, and continues until these
objectives are fully satisfied.
Step1 - Identification
● The identification stage of social media analytics involves finding the right sources of
data to analyze in order to gain valuable business insights.
● The data should be aligned with the business's objectives and can come from both
official business-owned platforms, such as social media accounts and blogs, and
nonofficial platforms such as Google search trends or Twitter search stream data.
● It is important to consider the business objectives when identifying the sources and
types of data to be analyzed.
Step2 - Extraction
● The extraction stage of social media analytics involves using appropriate methods and tools to
gather data from identified sources.
● The type (e.g., text, numerical, or network) and size of data will determine the method and tools
suitable for extraction.
● This can include manual data collection for small-scale data and automated extraction using APIs
(application programming interfaces) for larger data sets.
● It is important to consider privacy and ethical issues when mining data from social media
platforms and to have a clear social media privacy policy in place to ensure that data handling and
extraction practices do not violate user privacy.
● Specialized tools may be needed to extract certain types of data, such as social network and
hyperlink network data.
Step3 - Cleaning
● The cleaning step in social media analytics involves removing unwanted data from the
collected data set.
● This can involve processes such as cleaning, coding, filtering, clustering, and natural
language processing to remove irrelevant data.
● Both automated and manual techniques may be used for cleaning, depending on the
type of data and the desired level of accuracy.
Step 4- Analyzing
● The analyzing stage of social media analytics involves using clean data to identify
valuable insights for the business.
● The approach and techniques used will depend on the type of data being analyzed and
the tools and algorithms employed.
● It is important to maintain the integrity of the data while extracting meaningful
insights and to have a good understanding of the capabilities of the tools being used.
Step 5 - Visualization
● The visualization step in social media analytics involves creating visual representations of
the results of the analysis
● Visualization can help reveal hidden patterns, relationships, and trends in complex and large
data sets
● Different types of types of visualizations include charts, graphs, trees, heat maps, 3D views
● Effective visualization is important for effectively communicating the results of the analysis
to top management
● Text analytics can result in a word cooccurrence cloud; hyperlink analytics will provide visual
hyperlink networks; and location analytics can produce interactive maps.
● Depending on the type of data, different types of visualization are possible :
○ Network data (with whom)—network data visualizations can show who is connected to whom.
For example, a Twitter following-following network chart can show who is following whom.
○ Topical data (what)—topical data visualization is mostly focused on what aspect of a
phenomenon is under investigation. A text cloud generated from social media comments can
show what topics/themes are occurring more frequently in the discussion.
○ Temporal data (when)—temporal data visualization slice and dice data with respect to a time
horizon and can reveal longitudinal trends, patterns, and relationships hidden in the data.
○ Geospatial data (where)—geospatial data visualization is used to map and locate data, people,
and resources.
Step 6 : Interpretation
● Interpret and translate analytics results into a meaningful business problem.
● Two strategies or approaches used are:
1) Producing easily consumable analytical results
--training data scientists and analysts to produce interactive and easy-to-use
visual results
2) Improving analytics for understanding the consumption capabilities
Challenges to Social Media Analysis
● Volume and Velocity is a challenge
○ Social media data is large in size and is swiftly generated.
○ Capturing and analyzing millions of records that appear every second is a real challenge.
For example,
■ on Twitter, three-hundred forty-two thousand tweets appear every minute, and on
Facebook, one million likes are shared every twenty minutes.
○ Capturing all this information may not be feasible.
○ Knowing what to focus on is crucial for narrowing down the scope and size of the data.
○ Luckily, sophisticated tools are being developed to handle high-volume and high-velocity
data.
Challenges to Social Network Analysis
● Diversity Challenge- Social media users and content they generate are
extremely diverse, multilingual, and vary across time and space.
● Unstructuredness as a challenge
● Unlike the data stored in the corporate databases, which are mostly numbers, social
media data is highly unstructured and consists of text, graphics, actions, and relations.
● Short social media text, such as tweets and comments, has random grammatical
structure, and is laden with abbreviations, acronyms, and emoticons (a symbol or
combination of symbols used to convey emotional expressions in text messages), thus
representing a great challenge for extracting business intelligence.
Social Media Analytics Tools
Layer of social media Example of tools
Text Discovertext
Lexalytics Demo
:Video
Link
Archivist
Twitonomy
Netlytic
LIWC
Voyant
Actions Lithium Twitonomy
Google Analytics
Video Demo : Video
Link
SocialMediaMineR
Network NodeXL UCINET
Pajek Netminer
Flocker Netlytic
Reach
Mentionmapp
Mobile Countly
Mixpanel
Google Mobile Analytics
Google Fusion Table
Location Tweepsmap
Trendsmap
Followerwonk
Esri Maps Agos
Hyperlinks Webometrics Analyst VOSON
Research Engines Google Trends
Questions
● Why it is important for business managers to understand and mine social media
data?
● What is social media analytics, and how it is different from traditional business
analytics?
● Briefly explain the seven layers of social media data. Support your answer with
examples.
● Explain the social media analytics cycle.
● What ethical issues should be considered when mining social media data?
● What are some main challenges to social media analytics?
● Compare different social media analytics tools available in the market and explain
their strengths and weakness.