Download - Smarter Customer Analytics - Customer DNA
© 2012 IBM Corporation
“Know me” - Getting Closer to Your Customers through Applied AnalyticsMark Matiszik
© 2012 IBM Corporation
Know, Listen To, and Empower Me
Being treated as an
individual has moved
from a Desire to an
Expectation
© 2012 IBM Corporation
Case Study: How can “Customer Centricity” really help solve a business challenge?
Major US Retailer’s Big Unanswered Question:
“How do I eliminate unnecessary spend from my
Marketing budget?”
Step 1: Build a data-driven Lens of the Customer
Step 2: Apply that Customer Foundation as the key input
to Optimizing between Channels/Regions/Customers
© 2012 IBM Corporation
Balancing on a Thousand Curves
� The picture isn’t simple – there are many customers, and many media types.
� With two customer groups, for instance, we have two curves…
TV Spend
Customer
Spend
A
BC
D
E
1
2
3
Customer 1
Customer 2
But, how do we know what is optimal for each customer?
© 2012 IBM Corporation
Acting on Customer Insight
The customer’s voting record – the digital footprints of their countless decisions –
has the power to tell us who they are, and what matters to them.
Technology
Business Integration
Analytics
© 2012 IBM Corporation
The New Era of Customer Understanding and Segmentation
The key to achieving the high ROI and Profit potential of multi-channel shopping is
advanced customer analytics
Traditional Approach
� Models based on few dimensions
– demographics, value, or basket
You Are What You Buy
Latest & GreatestPrice FocusedValue MaximizersConnected Convenience
Customer Value
Typically not actionable because customers are more complex than 2 or 3 dimensions
Demographic
Trans-actions
Sales
Basket Analysis
Geo-graphy
Income
Advanced Clustering
� Models based on many dimensions
of customer behavior
Preferred ProductCategories
Preferred Channel
Participation in Loyalty Program
Use of In-House Credit Card
Use of Service Programs
Return/Exchange BehaviorBreadth of
Categories Shopped
Length of Timeas Customer
Recency + Frequency+ Value
Response to Media
Time until Repurchasein Key Categories
Highly actionable clusters are based on the customer’s response to various dimensions of the Retailer’s value proposition
© 2012 IBM Corporation
No Guessing: Analytics Can Reveal Who Your Customers are
� Begin with 30-40+ Modeled Variables from Customers’ Digital Footprints
� Each Variable is like a gene, which describes a facet of customer behavior
� Useful on their own, but also provide the input for Clustering
Age + Income + Geography
Preferred Product Categories
Modeled time to next purchase
CTP Customer
Use of In-House Credit Card
Facebook Page Engagement
Return / Exchange Behavior
Breadth of Categories Shopped
Length of Time as CustomerRecency + Frequency + Value
Response to Media
Gift Registry User
Annual Spend Level
Annual Transactions
Econometric: Real-estate & Unemployment
Most segmentation approachesonly focus here:
© 2012 IBM Corporation
Clusters are Based on the most significant Modeled Variables
Revolutionary customer segmentation approach tailored uniquely to each client’s business
model, customer data and operational practices, yielding highly actionable customer groups
Action Clusters are
Highly homogeneous – it is difficult to get into a cluster based on 10+ dimensions, ensuring that the customers are very similar to one another
Highly differentiated – the process ensures as much “distance” between clusters as possible
Preferred ProductCategories
Preferred Channel
Participation in Loyalty Program
Use of In-House Credit Card
Use of Service Programs
Return/Exchange BehaviorBreadth of
Categories Shopped
Length of Timeas Customer
Recency + Frequency+ Value
Response to Media
Time until Repurchasein Key Categories
© 2012 IBM Corporation
Sample Clusters
Rank Action Cluster % of Customers % of Spend
1 Brand Fanatics 9% 30%
2 Core Customers 8% 18%
3 Online Socialites 6% 14%
4 Hurt by the Economy 8% 8%
5 Potential Pool 7% 7%
6 Make it Interesting! 6% 6%
7 Let’s Bargain 10% 3%
8 Find me Online 7% 2%
9 Unengaged 17% 7%
10 Luxury for Me 4% 2%
11 Until Next Year (One and Done!) 13% 2%
12 Just Window Shopping 5% 1%
© 2012 IBM Corporation
This is the Outcome NExample: “Brand Fanatics”
Marketing Call to Action – RMI 37:1
EMOTIONAL BENEFIT: Sports enthusiast
BRAND PROMISE: Latest & Greatest, Multiple Sports Category Breadth and Depth
CUSTOMER AWARENESS: Loyalty promo, new product releases, direct mail and email
TOUCH POINTS: Multi-Channel, In-store and on web
UNIQUE IDEA: ‘Co-Branded Credit Card Promotion’
PRE-STORE: Mobile, Blogs, Social Networks
IN-STORE: Mobile applications and shopping aids, services merchandise together
POST-STORE: Online, loyalty program mailings and emails
Vital Statistics
� Strongest Loyalty: Over 85% are part of loyalty program
� 89% have shopped over 5 categories
� 91% have been customers for 7+ years
� Almost no new customers in <3 years
� 70% are due to purchase within 60 days
� 60% are using a private label credit card, 30% exclusively for all purchases
� Highest Return on Marketing scores
9% of customers ���� 30% revenue
© 2012 IBM Corporation
Clusters can deepen Insights from Existing Segmentations
© 2012 IBM Corporation
Technology
Next Step: Optimization
Use the lens of the Customer Foundation as a Primary Input for solving the
most difficult challenges within the business
Business Integration
Analytics
© 2012 IBM Corporation
Marketing Media Optimization – with a Customer Lens
• Multi-objectives
• Policy constraints
• Optimal decisions
Economy
Customer
Media
Analysis
Optimization
• Transaction data
• Modeled Variables
• Action clusters
• Industry data
• Systematic risk
• Demand forecast
• Performance data
• Saturation
• Action Exposures
Behaviors
© 2012 IBM Corporation
Benefits and Results
Case Study 1: Enterprise Marketing Media Mix Optimization
Solution
� Established customer foundation through
unique customer segmentation approach
� Enabled prescriptive media mix optimization engine for Marketing
investment against the Clusters
� Developed solution enabling ‘what if’
scenarios and returning ‘what’s best’ output
� Reduced saturation of budget 5-7% (~$50M)
� Outperformed industry with greatly reduced budget; identified $1B in additional revenue
� Improved conversion and engagement
Challenges & Background
� Optimally invest a $MMM+ advertising budget to maximize sales AND maintain/reduce market spend?
� How do I apply my knowledge of my
customers to determine the proper
proportion of investment in each marketing
type?
© 2012 IBM Corporation
Benefits and Results
Case Study 2: Re-engage ‘Lapsed Best Customers’ to Drive Revenue
Solution
� Developed Behavioral Models to enrich understanding of customers:
� Leveraged customer foundation to inform media preference, creative personality, and communication timing
� Selected customer list, designed 5 creative versions and delivered through preferred marketing channel
� ~100% and ~200% increased response rate over expected results, for two tested campaigns
� Reactivated customers drove $180/customer incremental revenue in 2 months
Challenges & Background
� How best to reactivate lapsed ‘best customers’ in loyalty program?
� Tailor copy, creative and offer based on customer preferences
� Minimize execution costs by identifying communication channels each customer would responsive most to
Response Rate
4,91%
11,69%10,43%
8,70%
0,59%3,80%
8,67%7,81%
6,59%
0,37%0%
2%
4%
6%
8%
10%
12%
28-Nov 5-Dec 12-Dec 19-Dec 26-Dec
Expected Response
Offer 2Offer 1
© 2012 IBM Corporation
Operationalizing Action Clusters Across the Organization
4/25/2012
What?Product
Who?Customer
How?Channel
When?Lifecycle Mgmt.
““““Who is in the market, what do they want and how do I inspire them to purchase?””””
““““What messages are relevant to my targeted customers and how and
when do I communicate with them?””””
Marketing
Merchants
© 2012 IBM Corporation
Technology
Final Step: Operationalize the Insights (and Repeat)
With the tools and capabilities, the organization begins to make better
decisions that no longer treat all customers alike.
Business Integration
Analytics
© 2012 IBM Corporation
Break-out session Round 2
Please join us during the break-out session
Get Personal!� Het belang van personalised promotions voor retailers
� Hoe kunt u dit realiseren binnen uw organisatie?
Mark Matiszik, Associate Partner, Retail Center of Competence, IBM
Ewald Hoppen, Team Lead Web Analytics / Senior Web Analyst, Wehkamp.nl
© 2012 IBM Corporation
Thank You!
Mark Matiszik
IBM Retail Center of Competence
@MarkMatiszik