9:
Predictive
Marketing
Anticipating Market Demand
with Proactive Action
Predictive Analytics in Marketing Forecast
1 Who will buy
What will sell
Which campaign will work
01
Historical Data
02
Statistical algorithms Customer Behavior
2 When will they buy
Growth in ticket size
03
Machine learning
04 Optimization
Artificial Intelligence 3 Filtering ideas for NPD
Better ROI
Discovering ‘black box’
Predictive Marketing Applications
Customer Management – Customer Equity Model
Improved Customer Targeting - CLV
Predictive models help identify high-potential customers and tailor marketing campaigns
to their needs.
Enhanced Customer Retention - NBA
By predicting churn, businesses can implement retention strategies to keep customers
engaged across multiple channels from pre-sales to post-sales.
Optimized Marketing Spend and Customer Tiering
Predictive analytics allows for more efficient allocation of marketing budgets b/w high-
impact areas – acquisition or service - and high potential customers
Predictive Product Management
Launch Market Testing
Saves significant Product design and
part of the development prototype having higher
costs chance of success
Competitive edge Ideation
Proactively leverages Combination of right
external information on features basis existing
what is trending product analysis
Recommendations Personalization
Which product to Propensity model estimates
offer from an existing the likelihood of customers
portfolio – product with specific profiles to buy
affinity analysis certain products
Predictive Brand Management
Campaign Optimization
Mapping the key ingredients of successful past
campaigns
Sentiment analysis Jupiter
How our customers feel about our brands
Jupiter is the biggest planet
and campaigns in social media and third-
of them all
party review websites
CTR Analysis Dynamic website
Which content & creatives drive Easily track the customer
the most clicks journey across multiple websites and
social media to show next best
Audience–content fit content
Guide content distribution to the
right audience at the right time
Building Predictive Marketing Models
Regression Modeling
For simple predictions correlating two variables.
Best for straightforward, structured problems.
Collaborative Filtering
For recommendation systems based on
collaboration of customers to rate
products for the model to work
Neural Network
For complex unstructured predictions
based on how neural network operates
inside the human brain
Framework for Implementing Predictive Marketing
1 3 5
Data Collection Model Actionable
and Integration Building Insights
2 4 6
Data Cleaning and Continuous
Model Deployment
Preprocessing Improvement
and Monitoring
Data Collection
1 Gather data for both dependent and
independent variables
Equation Formation
2 Use statistical software to create an equation
explaining the variable relationships
Regression
Modeling Interpretation
Earth
3 Analyze the equation for insights
accuracy.
and
Prediction
Applications: Customer equity, propensity, churn
detection, and product affinity models.
4 Use the equation to predict outcomes based on
new independent variables
Collaborative Filtering
Preference Collection
Gather customer
preferences through
ratings or actions
Clustering
Group similar customers
or products based on
ratings and behaviors.
Rating Prediction
Predict ratings for new
Applications: Product and content products using data from like-
recommendations minded customers
Neural Network for Complex Predictions
Applications: Suitable for unstructured data and unknown algorithms.
Connection Discovery
Neural networks find
patterns and relationships in
the data
Data Loading Output Prediction
Input independent Use the discovered model to predict
variables and output outcomes from new inputs, refining
dependent variables into accuracy over time
the network
Summary
10:
Contextual
Marketing
Making a Personalized Sense-and-
Respond Experience
Contextual Marketing : Making a Personalised Sense-
and-Respond Experience
Once the infrastructure
It leverages data about a is in place, marketers will
customer’s behavior, only need to define the
preferences, location, and triggers and
other contextual factors responding actions.
to create highly relevant
marketing experiences.
Contextual marketing is The most critical element
When the right person
about delivering the right to establish AI-powered
with the right profile is
message to the right contextual
nearby the sensors,
person at the right time, marketing is to build a
marketers can learn more
based on their current connected ecosystem of
about the person and
context. sensors and devices,
offer the right products
either in the POS or on
with the right message.
customer premises.
Contextual Marketing Mechanism
Using Sensors for Contextual Response
The first step to create AI- Wearable devices such as
powered contextual smartwatches, fitness bands
marketing is to set up
a connected ecosystem of 1 6 can be useful devices in
future
sensors and devices,
especially at the
point of sale (POS)
marketers need to leverage
the device that is
2 5
Beacon – a Bluetooth low always in the customer's
energy transmitter is widely possession as a proxy for the
used to track customer whereabouts.
location A smartphone is one
alternative
3 4
The sensors can also help The challenge is to recognize the
marketers send personalized identity or the profile of the
content to the connected customer to ensure that the
devices, for response is truly personalized
instance, in the form of push
notifications.
Biometrics to Trigger Personalised Actions
Facial Voice Eye Tracking
Neuro-Signals
Recognition Recognition Sensor
• Demographic • AI can analyze the • Companies can • British Airways launched
profiles, individual properties of vocal understand where a the “happiness blanket,”
persons speech—speed, brief viewer focuses which can change color
• Disney used this to pauses, and tones— attention based on based on a passenger's
detect emotions and discover eye movements, for state of mind. The
during films embedded emotions example, when blanket came with a
• Automakers are seeing an ad or a headband that monitors
testing to enhance video brain waves
driving experience
Creating a Direct Channel to Customer Premises
Hershey and 3D System’s CocoJet can
print chocolate of various shapes and put a
personalized message on
a chocolate bar
Smart speakers such as Amazon Echo, Google
Nest, and Apple HomePod are powered by
intelligent voice assistants : Alexa, Google
Assistant, and Siri
The intelligent appliance ecosystem empowers
marketers to be instantly available with the right
products and services the moment customers need
them the most
For the Tide brand, P&G created an Alexa skill that
answers hundreds of questions about laundry.
Campbell's Alexa skill provides answers to recipe
inquiries.
For B2B, the companies can provide the contextual
data to their customers for preventive maintenance
regularly and eventually save costs
Samsung's Family Hub—a refrigerator with a
touchscreen display— allows shoppers to build a
shopping list and order groceries directly from the
Instacart app
1. Personalised Information
• Targeted messages are broadcasted to the
audience using geofencing through GPS
data
• Companies can use geofencing to drive
traffic to their stores from nearby locations
or competitors’ locations with promotional
offers
Three Levels of 2. Customized
Personalised Interaction
• Customers are engaged in a dialoge
• Companies trigger customers to move
Experience from awareness stage to action stage
through right incentives and offers
3. Total Immersion
• Using AR and robotics to provide a complete
experience in the physical space
• Fashion brands, such as Ralph Lauren, are using
the smart fitting room to provide immersive
digital experience in the physical world
Triggers and Responses in Contextual Marketing
