7. 7
“ When [AI] finally works, it
gets co-opted by some other
part of the field. So, by
definition, no AI ever works;
if it works, it’s not AI.
Source: https://ocw.mit.edu/courses/electrical-engineering-and-computer-science/6-825-
techniques-in-artificial-intelligence-sma-5504-fall-2002/lecture-notes/Lecture1Final.pdf
“
9. • Mitigation of compliance risks
• Removal of Personally Identifiable
Information (PII)
• Removal of Redundant, Outdated and
Trivial (ROT) content
• Protection of intellectual property
• Identification of patterns of fraud
• Detection of customer sentiment
• Prediction of credit risks
• Feature extraction from product data
Text Analytics,
a long time staple
of content
management, is
now called “AI”
20. PRODUCT ARCHITECTURE
COMPONENTIZED CONTENT
CUSTOMER DATA FOUNDATION
DELIVERING PERSONALIZED CUSTOMER EXPERIENCES AT SCALE
What does it take to do this right?
advanced services & expertise
unified models standardized platforms & processes
20
• enriched customer journeys
• product attribute model & corresponding
taxonomies
• data intake, clean-up, aggregation.
• analysis, recommendation & decision
making
(ML, data science, human judgment)
• process setup (continuous or periodic)
• standard pipeline for insight
delivery to marketing teams
KNOWLEDGE & INSIGHTS
• product data with e-catalog and display
hierarchies optimized for customer
journeys
• back end product information onboarding
process aligned with customer
experience practices
• metrics driven decision making
• merchandizer collaboration with product
and solution experts
• configure price quote and recommendation
tools aligned with user personas and pain
points
• product information management
ecosystem aligned with rich media
• cross sell and upsell relationships
• merchandizing and solution bundles
• optimized content structure
• component architecture aligned with
messaging architecture
• content attribute model & corresponding
taxonomies
• omnichannel offer recommender
• dynamic offer generator
• content assembly based on offering
architecture and baseline hypotheses
tested against target outcomes
• recombination tested continuously using
changing messaging architecture
• component content management
system
• content production workflows
• content standards & governance
• high fidelity customer journeys with
augmentation and automation
opportunities
• customer attribute model &
corresponding multi-dimensional
audience taxonomies
• profile standardization
• pattern recognition
• customer signal reconciliation across
upstream platforms
• machine learning development & training
• customer data platform
• customer data modeling
• cross system normalization
• metrics aligned data governance
decision making