PREDICTIVE
ANALYTICSThe power to predict
who will click, buy, lie, or die
Why Use Predictive Analytics in Business
DECISION
Why use predictive analytics in business? right business decision —> — customer behavior, — dealing with overwhelming complexity, — hundreds or even thousands of factors, — a universe of thousands or millions of customers,— people just cannot “connect the dots” to make the ideal decision,— ‘Predictive Analytics’ connects the dots scientifically, guiding each decision to greater success.
Google: Predicts which new ads will get many bounces (when people click on an ad, but then immediately click the back button).
Stanford University: Derived with predictive modelling, an innovative method that diagnoses breast cancer better than human doctors in part by considering a greater number of factors in a tissue sample.
Example organizations that use predictive analytics
Example organizations that use predictive analytics
• Obama was re-elected in 2012 with the help of voter prediction.
• The leading career-focused social network, LinkedIn, predicts your job skills.
• Target Co. predicts customer pregnancy in order to market relevant products accordingly
• Online dating leaders Match.com, OkCupid, Tinder predict which hottie on your screen would be the best bet on your side
Example organizations that use predictive analytics
Hewlett-Packard (HP) earmarks each and every one of its more than 330,000 worldwide employees according to “Flight Risk” the expected chance he or she will quit their job, so that managers may intervene in advance where possible and plan accordingly otherwise.
Inspired by the TV crime drama Lie to Me about a micro-expression reader, researchers at the University at Buffalo trained a system to detect lies with 82 percent accuracy by observing eye movements alone.
Computers can literally read your mind.
Example organizations that use predictive analytics
amazon.com: 35 percent of sales come from product recommendations.
Netflix: Sponsored a $1 million competition to improve movie recommendations; a reported 70 percent of Netflix movie choices arise from its online recommendations
Target: Increased revenue 15 to 20 percent by targeting direct mail with product choice models.
BIG DATA
Half the money I spend on advertising is wasted; the trouble is I don’t know which half.
John Wanamaker
“ “If you torture the data long enough, it will confess.
RONALD COASE, Professor of Economics, University of Chicago
“ “
Technology that learns from experience (data) to predict the future behaviour of individuals in
order to drive better decisions.
Insider Predictive Analytics
Insider Predictive Analytics
Predictive analytics simplifies data to amplify value
Predictive analytics can navigate overwhelming complexity to give you a clearer view of the future.
PAST PRESENT FUTURE
Data Predictive Analytics Uncertainty Future
Outcomes
Which action will drive the best
result
Data Integration
& Optimization
Advanced Analytics
What is likely to happen in the
future
Analyze the Customers’ Behaviors
Predictive Analytics
Campaign Execution
Predicted Customer Insights
Auto Optimization
Information from business systems
Optimal Action
Analytics Reporting
Real-time Updates
Advanced JS Mobil SDK Surveys Offers
Partner Website
Auto Segmentation
Predict a visitor’s likely next move and act on real time via our machine learning platform.
Predictive SegmentationLikelyhood yerine likelihood olacak
WHICH BEHAVIORAL SEGMENTS PRODUCE MOST REVENUE? DISCOUNT SHOPPERS
LIKELYHOOD TO PURCHASE
AT LEAST PURCHASED ONCE
SLEEPLESS
LUXURY SHOPPERS
WINDOW SHOPPERS
NOT PURCHASED FOR 6 MOUTHS
RECENT VISITORS
NEVER PUCHASED
BEHAVIORAL SEGMENTS NEVER PUCHASED
NOT PURCHASED FOR 6 MONTHS
AT LEAST PURCHASED ONCE
DISCOUNT SHOPPERS
LUXURY SHOPPERS
WINDOW SHOPPERS
LIKELYHOOD TO PURCHASE
RECENT VISITORS
SLEEPLESS
If the visitor comes to website from Facebook
Rule Model : Likelihood to Purchase
Characteristic of and individual
Predictive Model
Predictive Score
has searched on website more than
3 times
average spent time on product detail pages
between 120 seconds and 480 seconds
has checked an item more than once
has been less than 20 days since his/her last
purchase
THEN the probability of purchasing at the end of this session is %75.
& & & &
Let algorithms discover your customer segments which acquired best performance
Auto Segmentation
Segment Categories
o Demographic Segments
o Behavioral Segments
o Interest Segments
$124 average order $595 total revenues 67 days between orders 5 total orders
14 total items $164 first order revenues 3.3 products in first order 3% of orders on clearance +10 more
High value, fewer orders, Big spend on 1st order
$99 average order $2.261 total revenues 24 days between orders 24 total orders
57 total items $76 first order revenues 1.7 products in first order 6% of orders on clearance +10 more
Long term, high value, frequent buyers
Example : Cluster DNA
Demographic Segments
Gender
Male Female
Age
18-24 25-34 35-45 Over 45
Income
High Medium Low
Marital Status
Married Not married
Churn / Loyal Customers
Never purchased
Purchased at least once
Frequent buyers
Non recent visitors
Sleepless
Discount Shoppers
Window Shoppers
Luxury Shoppers
Likelihood to purchase (High, medium, low)
Spending Pattern (Beginning of month, end of month etc.)
Behavioral Segments
Interest Segments
Cluster X Brand Scale Cluster Y Brand Scale
Preferred BrandsTahari Arthur S.Levine
Adrienne Vittadini Calvin Klein
Eliza J Nine West
Preferred BrandsDesigual
Dzhavael Couture Custo Barcelona
Smash! Salvage
Least InterestCollective Concepts
Wow Couture Max and Cleo
Rene Rofe Steven
Least InterestPleasure Doing Business
Wow Coutre Desigual
6126 L*Space
Example : Brand Clusters DNA